Task execution method, task execution device and electronic equipment

Abstract

The invention discloses a task execution method, a task execution device and electronic equipment, and belongs to the technical field of artificial intelligence. The interaction method comprises the steps that under the condition that task request information input by a user is displayed in a dialogue input box of an assistant interface, at least one task label is displayed on the assistant interface; wherein the assistant interface is an interface of an artificial intelligence AI assistant, one task label is used for indicating one piece of prediction information generated according to the task request information, and the prediction information comprises task elements missing from the first task; under the condition that the first input is received, responding to the first input, and displaying a first dialogue message in a dialogue information display area of the assistant interface; the first dialogue message comprises task content describing the first task, and the task content comprises all task elements required for executing the first task; and executing the first task in response to the second input under the condition that the second input is received.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence technology, and specifically relates to a task execution method, a task execution device, and an electronic device. Background Art

[0002] In related technologies, electronic devices typically feature artificial intelligence (AI) assistants, allowing users to interact with them to complete desired operations. For example, users can interact with AI assistants to book tickets, search for information, and so on. However, in practice, users often need to engage in multiple rounds of dialogue and interaction with AI assistants to complete operations such as booking tickets and searching for information. This requires excessive user input, cumbersome operation steps, and the model needs to parse multiple levels of instructions, which can easily lead to misunderstandings of user intent and reduce the efficiency of the interaction process. Summary of the Invention

[0003] The purpose of the embodiments of the present application is to provide a task execution method, a task execution device, an electronic device and a readable storage medium, which can solve the problem of low efficiency of the interaction process when users interact with artificial intelligence (AI) assistants.

[0004] In a first aspect, an embodiment of the present application provides a task execution method, the task execution method comprising:

[0005] When the task request information input by the user is displayed in the dialogue input box of the assistant interface, at least one task label is displayed on the assistant interface; wherein the assistant interface is an interface of the artificial intelligence (AI) assistant, and a task label is used to indicate a piece of prediction information generated based on the task request information, and the prediction information includes the task elements missing from the first task;

[0006] Upon receiving the first input, in response to the first input, displaying a first dialogue message in the dialogue information display area of the assistant interface; the first dialogue message comprises task request information and tag content of a task tag selected by the user determined based on the first input; the first dialogue message comprises task content describing the first task, and the task content comprises all task elements required to perform the first task;

[0007] In a case where the second input is received, the first task is performed in response to the second input.

[0008] In a second aspect, an embodiment of the present application provides a task execution method, including:

[0009] When the third conversation message is displayed in the conversation information display area of the assistant interface, at least one task card is displayed on the assistant interface; wherein the assistant interface is an interface of an artificial intelligence (AI) assistant, and each task card includes prediction information generated based on the third conversation message, and the prediction information includes missing task elements of the third task; each task card includes task trigger information determined based on the third conversation message and historical reference information; the task trigger information includes all task elements required for the second application to perform the third task and the element content of each task element; the third conversation message includes task content describing the third task; and the element content of each task element is the element content with the highest prediction accuracy.

[0010] receiving a first input from a user regarding at least one task card;

[0011] In response to the first input, a third task is performed.

[0012] In a third aspect, an embodiment of the present application provides a task execution device, including:

[0013] A display module, configured to display at least one task tag on the assistant interface when the task request information input by the user is displayed in the dialogue input box of the assistant interface; wherein the assistant interface is an interface of an artificial intelligence (AI) assistant, and a task tag is used to indicate a piece of prediction information generated based on the task request information, and the prediction information includes the task elements missing from the first task;

[0014] Upon receiving the first input, in response to the first input, displaying a first dialogue message in the dialogue information display area of the assistant interface; the first dialogue message comprises task request information and tag content of a task tag selected by the user determined based on the first input; the first dialogue message comprises task content describing the first task, and the task content comprises all task elements for executing the first task;

[0015] The execution module is configured to execute the first task in response to the second input when the second input is received.

[0016] In a fourth aspect, an embodiment of the present application provides a task execution device, including:

[0017] A display module is configured to display at least one task card on the assistant interface when the third dialogue message is displayed in the dialogue information display area of the assistant interface; wherein the assistant interface is an interface of an artificial intelligence (AI) assistant, and each task card includes prediction information generated based on the third dialogue message, the prediction information including missing task elements of the third task; each task card also includes task trigger information determined based on the third dialogue message and historical reference information; the task trigger information includes all task elements required for the second application to perform the third task and the element content of each task element; the third dialogue message includes task content describing the third task; the element content of each task element is the element content with the highest prediction accuracy;

[0018] A receiving module, configured to receive a first input of a user on at least one task card;

[0019] The execution module is configured to execute a third task in response to the first input.

[0020] In a fifth aspect, an embodiment of the present application provides an electronic device comprising a processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the programs or instructions are executed by the processor, the steps of the method of the first aspect or the second aspect are implemented.

[0021] In a sixth aspect, an embodiment of the present application provides a readable storage medium, which stores a program or instruction. When the program or instruction is executed by a processor, the steps of the method of the first aspect or the second aspect are implemented.

[0022] In the seventh aspect, an embodiment of the present application provides a chip, which includes a processor and a communication interface, the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the steps of the method of the first aspect or the second aspect.

[0023] In an eighth aspect, an embodiment of the present application provides a computer program product, which is stored in a readable storage medium and is executed by at least one processor to implement the method of the first aspect or the second aspect.

[0024] The task execution method provided by the embodiment of the present application can predict the actual task elements of the first task based on the task request information when the user enters the task request information in the dialogue input box of the assistant interface, that is, it realizes the prediction of the user's true intention. Furthermore, at least one task label is displayed in the assistant interface to provide the user with at least one result that the user may need for the user to select. When the user's first input is received, the first dialogue message can be displayed in the dialogue information display area to realize the display of all task elements required to execute the first task. Furthermore, when the user's second input is received, the first task that the user needs to execute can be executed. During the execution of the first task, the user only needs to enter the task request message in the dialogue input box, and the artificial intelligence AI assistant can automatically predict all task elements of the first task that the user needs to execute, without the user having to input more detailed information or multiple rounds of information, thereby effectively improving the execution efficiency of the electronic device for the first task that the user needs to execute. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A flowchart of a task execution method provided for some embodiments of the present application;

[0026] Figure 2 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0027] Figure 3 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0028] Figure 4 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0029] Figure 5 A flowchart of a task execution method provided for some embodiments of the present application;

[0030] Figure 6 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0031] Figure 7 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0032] Figure 8 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0033] Figure 9 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0034] Figure 10A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0035] Figure 11 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0036] Figure 12 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0037] Figure 13 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0038] Figure 14 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0039] Figure 15 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application;

[0040] Figure 16 A flowchart of a task execution method provided for some embodiments of the present application;

[0041] Figure 17 A structural block diagram of a task execution device provided in some embodiments of the present application;

[0042] Figure 18 A structural block diagram of a task execution device provided in some embodiments of the present application;

[0043] Figure 19 A structural block diagram of an electronic device provided for some embodiments of the present application;

[0044] Figure 20 A schematic diagram of the hardware structure of an electronic device provided for implementing some embodiments of the present application. DETAILED DESCRIPTION

[0045] The following will be combined with the accompanying drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the embodiments described are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of this application.

[0046] The terms "first," "second," and the like in the specification and claims of this application are used to distinguish similar objects and do not include descriptions of a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate so that the embodiments of this application can be implemented in an order other than that illustrated or described herein, and that the objects distinguished by "first," "second," and the like are generally of the same type and do not limit the number of objects; for example, the first object can be one or more. In addition, the term "and / or" in the specification and claims indicates at least one of the connected objects, and the character " / " generally indicates that the objects connected are in an "or" relationship.

[0047] The terms used in the embodiments of this application are only used to explain the specific embodiments of this application and are not intended to limit this application. The following is an explanation of the terms involved in the embodiments of this application.

[0048] Interface: The user interface (UI) is the media interface for interaction and information exchange between applications or operating systems and users. It converts the internal form of information into a form acceptable to the user. The user interface is source code written in a specific computer language such as Java or Extensible Markup Language (XML). The interface source code is parsed and rendered on the electronic device, ultimately presenting content that the user can recognize. The most common form of user interface is the graphical user interface (GUI), which refers to a user interface related to computer operations that is displayed graphically. It can be visual interface elements such as text, icons, buttons, menus, tabs, text boxes, dialog boxes, status bars, navigation bars, widgets, etc. displayed on the display screen of an electronic device.

[0049] Artificial Intelligence (AI) Assistant: An embedded system service or functional module with natural language processing and machine learning capabilities for human-computer interaction via voice or text interfaces. This assistant is an intelligent assistance tool based on AI technology that can automate tasks, provide search and information services, personalize responses, and optimize the user experience. It aims to help users efficiently complete various tasks, obtain information, or create content. The AI assistant has natural language processing capabilities and can understand and process natural language commands input by users via voice or text. This includes recognizing, parsing, and generating appropriate responses. Users can engage in conversations with the AI assistant through its assistant interface. Conversation messages between the user and the AI assistant refer to the information exchanged during the interaction. These messages encompass both user input and the AI assistant's responses. Specifically, user conversation messages can include queries or commands—requests or commands issued to the AI assistant via typing on an input method keyboard or voice input, such as asking about the weather, requesting to set a reminder, or seeking information. Conversation messages from the AI assistant can also include responses—the answers or results of actions generated by the AI assistant in response to user queries or commands.

[0050] Model: refers to the neural network model, which is a complex network system formed by a large number of simple processing units, that is, neurons that are widely interconnected. It reflects many basic characteristics of human brain functions and is a highly complex nonlinear dynamic learning system.

[0051] Slots: In conversational systems, slots are variable spaces used to extract specific information from user input. For example, in a ticket booking conversation, slots can be used to capture information such as date, destination, and departure point. Filling in slots is a key step in understanding user intent and extracting information to help the system perform appropriate actions or responses.

[0052] Slot parameters: In a dialogue system, these refer to semantic data extracted from user statements, used to fill slots and support intent execution. Parameters are captured dynamically during the dialogue process, meaning they may be gradually filled or modified as the conversation progresses. For example, if a user says, "Book a ticket to location A on Friday," "Friday" and "location A" serve as parameters for the time and location slots, driving the system to call the ticket booking API.

[0053] Task elements: refers to the basic components or conditions required to complete a task. For example, the task elements of a navigation task include the departure place, destination, mode of travel, etc. The task elements of a ticket booking task include the departure place, destination, travel time, type of transportation, etc.

[0054] It should be noted that the task execution method provided in the embodiments of the present application can be executed by electronic devices such as mobile phones, tablets, laptops, PDAs, and in-vehicle electronic devices. In some embodiments of the present application, the task execution method provided in the embodiments of the present application is executed by an electronic device as an example to illustrate the task execution method. One specific application scenario is a medical consultation task, in which the user can interact with an artificial intelligence AI assistant, which obtains preliminary symptom information input by the user. The artificial intelligence AI assistant automatically predicts and completes the accompanying symptom options based on the preliminary symptom information, further loads detection suggestions, and pushes them to the user. Another specific application scenario is a navigation task, in which the user can interact with an artificial intelligence AI assistant and input the type of destination, such as a charging station, a gas station, a shopping mall, etc. The artificial intelligence AI assistant predicts the specific destination location for the user based on the remaining mileage of the vehicle, the user's preference for the type of destination, etc., and further recommends a navigation application to the user to complete the navigation task.

[0055] The task execution method, task execution device, and electronic device provided in the embodiments of the present application are described in detail below with reference to specific embodiments and their application scenarios in conjunction with the accompanying drawings.

[0056] In some embodiments of the present application, a task execution method is provided. Figure 1 Flowchart of a task execution method provided for some embodiments of the present application; Figure 1 As shown, the task execution method is executed by the electronic device, and the task execution method includes:

[0057] Step 102: When the task request information input by the user is displayed in the dialogue input box of the assistant interface, at least one task label is displayed on the assistant interface; wherein the assistant interface is an interface of the artificial intelligence (AI) assistant, and a task label is used to indicate a piece of prediction information generated based on the task request information, where the prediction information includes the missing task elements of the first task;

[0058] In some embodiments of the present application, Figure 2 This is one of the interface diagrams of the task execution method of the embodiment of the present application, such as Figure 2As shown, when the user interacts with the artificial intelligence AI assistant, the user can input task request information in the dialogue input box 202 of the artificial intelligence AI assistant. Among them, the task request information is the information required by the user for the task that the artificial intelligence AI assistant wants to perform. At this time, the artificial intelligence AI assistant can make predictions based on the task request information input by the user and generate at least one task tag, and then display the at least one task tag in the assistant interface of the artificial intelligence AI assistant. Among them, the task tag can be used to indicate the movement prediction information generated according to the task request information, and the prediction information can include the missing task elements of the first task that the user needs to perform. That is, all task elements of the first task that the user needs to perform are predicted.

[0059] For example, Figure 2 A schematic diagram of an interface of a task execution method provided in some embodiments of the present application; Figure 3 This is a schematic diagram of an interface of a task execution method provided in some embodiments of the present application, such as Figure 2 and Figure 3 As shown, when a user interacts with an AI assistant, they enter the task request "Book me a ticket" into the AI assistant's dialogue input box 202. The AI assistant then makes a prediction based on the "Book me a ticket" message and determines at least one item of prediction information. The prediction information is information necessary for the task the user may need to perform, as predicted by the AI assistant based on the task request input. The AI assistant then generates and displays three task labels in the assistant interface: task label 204, task label 206, and task label 208, representing "Airplane ticket to location A," "High-speed rail to location A," or "Ticket to scenic spot B." The content displayed in the three task labels, "Airplane ticket to location A," "High-speed rail to location A," or "Ticket to scenic spot B," represents the missing task elements of the first task. At this point, the AI assistant predicts that the first task the user needs to perform is "Book a plane ticket to location A," "Book a high-speed rail to location A," or "Book a ticket to scenic spot B."

[0060] In addition, the user inputs task request information in the dialogue input box of the assistant interface, enters text in the dialogue input box through a touch device such as a finger or a stylus, or obtains the user's voice input, extracts the voice information entered by the user, and generates task request information in the dialogue input box. The specific details can be determined according to actual usage requirements and are not limited in the embodiments of this application.

[0061] Step 104, when the first input is received, in response to the first input, a first dialogue message is displayed in the dialogue information display area of the assistant interface; the first dialogue message is composed of task request information and the label content of the task label selected by the user determined based on the first input; the first dialogue message includes the task content describing the first task, and the task content includes all task elements required to perform the first task. The first input can be the user's input to any task label or the user's input to the dialogue input box or input method keyboard in the assistant interface. The first input is used to select a task label. The above-mentioned first input includes but is not limited to: the user's touch input to any task label, the dialogue input box or input method keyboard in the assistant interface through a touch device such as a finger or a stylus, or a voice command input by the user, or a specific gesture input by the user, or other feasible input, which can be determined according to actual usage needs and is not limited in the embodiments of this application. The specific gesture in the embodiment of the present application can be any one of a single-click gesture, a sliding gesture, a drag gesture, a pressure recognition gesture, a long press gesture, an area change gesture, a double-press gesture, and a double-click gesture; the click input in the embodiment of the present application can be a single-click input, a double-click input, or any number of click inputs, etc., and can also be a long press input or a short press input.

[0062] Upon receiving the user's first input, the AI assistant can display the first conversation message in the conversation information display area of the assistant interface. It is understood that the conversation information display area is different from the conversation input box. The information in the conversation input box is the information that the user is currently entering during the interaction with the AI assistant but has not yet been displayed in the conversation information display area. The information in the conversation information display area is the information that the user confirms to send or that the AI assistant automatically sends to the screen.

[0063] The first conversation message is composed of the task request information entered by the user in the conversation input box and the tag content of the task tag selected by the first input. In other words, the task request information entered by the user and the tag content of the task tag are concatenated to form the complete first conversation message. The first conversation message includes the task content of the first task, which includes all task elements required to execute the first task.

[0064] For example, Figure 4 Schematic diagram of the interface of the task execution method provided in some embodiments of the present application; Figure 3 and Figure 4As shown, after the user makes the first input, it is first determined that the task label selected by the user's first input is "Air ticket to location A", and the task request information entered by the user in the dialogue input box is "Book a ticket for me". At this time, the artificial intelligence AI assistant can splice "Book a ticket for me" and "Air ticket to location A" to generate a first dialogue message 220, that is, "Book a ticket for me to location A", and display it in the dialogue information display area.

[0065] Step 106 : When the second input is received, execute the first task in response to the second input.

[0066] In some embodiments of the present application, upon receiving a first input from a user, the AI assistant may display a first dialogue message in the information display area of the assistant interface. The first dialogue message includes a description of the first task, and the task content includes all task elements required to perform the first task, indicating that all conditions for performing the first task have been met. At this point, the AI assistant may receive a second input from the user. Upon receiving the second input from the user, the AI assistant may execute the first task in response to the second input, thereby completing the task execution process.

[0067] It is understandable that during the execution of the first task, the user only needs to enter the task request information in the dialogue input box of the assistant interface. The artificial intelligence AI assistant can predict all the task elements of the first task that the user needs to perform based on the task request information, and automatically provide the user with the application required to perform the first task, thereby completing the execution of the first task. There is no need for the user to input more detailed information or multiple rounds of information input, which can effectively improve the efficiency of the electronic device in executing the first task that the user needs to perform. Taking the ticket booking task as an example, the traditional ticket booking process through the artificial intelligence AI assistant may require at least 4 to 5 rounds of input to complete the ticket booking, while the ticket booking process of this application only requires 2 to 3 rounds of interaction to complete the ticket booking, which greatly improves the efficiency of executing the ticket booking task.

[0068] For example, Figure 4As shown, after the user makes the first input, it is first determined that the task label selected by the user's first input is "ticket to location A", and the task request information entered by the user in the dialogue input box is "book me a ticket". At this time, the artificial intelligence AI assistant can splice "book me a ticket" and "ticket to location A" to generate a first dialogue message 220, that is, "book me a ticket to location A", and display it in the dialogue information display area. At this time, the artificial intelligence AI assistant can call the corresponding application for booking tickets, and when it receives the user's second input, it can run the application to complete the task of booking a ticket to location A. The above-mentioned second input includes but is not limited to: the user's touch input of any interface element in the assistant interface used to perform the first task through a touch device such as a finger or a stylus, or a voice command entered by the user, or a specific gesture entered by the user, or other feasible input. The specific input can be determined according to actual usage needs and is not limited in the embodiments of this application. The specific gesture in the embodiment of the present application can be any one of a single-click gesture, a sliding gesture, a drag gesture, a pressure recognition gesture, a long press gesture, an area change gesture, a double-press gesture, and a double-click gesture; the click input in the embodiment of the present application can be a single-click input, a double-click input, or any number of click inputs, etc., and can also be a long press input or a short press input.

[0069] The task execution method provided by the embodiment of the present application can predict the actual task elements of the first task based on the task request information when the user enters the task request information in the dialogue input box of the assistant interface, that is, it realizes the prediction of the user's true intention. Furthermore, at least one task label is displayed in the assistant interface to provide the user with at least one result that the user may need for the user to select. When the user's first input is received, the first dialogue message can be displayed in the dialogue information display area to realize the display of all task elements required to execute the first task. Furthermore, when the user's second input is received, the first task that the user needs to execute can be executed. During the execution of the first task, the user only needs to enter the task request message in the dialogue input box, and the artificial intelligence AI assistant can automatically predict all task elements of the first task that the user needs to execute, without the user having to input more detailed information or multiple rounds of information, thereby effectively improving the execution efficiency of the electronic device for the first task that the user needs to execute.

[0070] In some embodiments of the present application, before displaying at least one task label on the assistant interface, the following steps are further included:

[0071] Inputting the task request information into the intention prediction model, and the intention prediction model generating at least one item of prediction information of the first task based on the task request information;

[0072] The intention prediction model determines a first application for performing the first task based on the task request information and the at least one prediction information;

[0073] The intention prediction model determines, based on the at least one prediction information and the first application, at least one slot required to perform the first task and slot parameters of each slot;

[0074] The intent prediction model generates at least one task label according to the first application, at least one slot, and slot parameters of each slot.

[0075] In some embodiments of the present application, when a user enters task request information in the dialogue input box of the assistant interface, the artificial intelligence AI assistant can use the intent prediction model to predict at least one item of prediction information of the first task that the user needs to perform, and use the intent prediction model to determine the first application required to perform the first task.

[0076] Specifically, after the user enters the task request information in the dialogue input box of the assistant interface, the artificial intelligence AI assistant can first input the task request information into the intention prediction model. The intention prediction model can process the task request information through its own neurons based on the task request information, thereby generating at least one item of prediction information for the first task, that is, predicting the missing task elements of the first task that the user needs to perform.

[0077] For example, the user enters "help me book a ticket" in the dialogue input box of the artificial intelligence AI assistant. At this time, the artificial intelligence AI assistant can input the information "help me book a ticket" into the intention prediction model. Based on this information, the intention prediction model determines that the user may need to book a plane ticket or train ticket to a certain place, or a ticket to a certain scenic spot. Here, a certain place, plane ticket, train ticket, a certain scenic spot, and ticket are the missing task elements of the first task, that is, the prediction information.

[0078] Furthermore, after determining at least one item of prediction information, the intention prediction model may also determine a first application for executing the first task based on the task request information and the at least one item of prediction information.

[0079] Taking the ticket booking task as an example, when the intention prediction model predicts that the first task the user needs to perform is the ticket booking task, it can search for programs that can be used for ticket booking from all the applications installed in the electronic device, such as map programs, travel programs, etc., that is, determine the first application required to perform the first task.

[0080] Furthermore, after determining the first application required to execute the first task, the intention prediction model can also determine at least one slot required to execute the first task and slot parameters of each slot based on the prediction information and the corresponding first application.

[0081] Taking the ticket booking task as an example, a ticket booking task requires at least the following slots to satisfy all the task elements: departure location, destination, travel method, and travel time. Accordingly, the corresponding relationship between the slots and their corresponding slot parameters is: departure location: location C, destination: location A, travel method: flight, travel time: March 18th.

[0082] Furthermore, after determining the first application, at least one slot and the slot parameters of each slot, the intention prediction model can generate at least one task label based on the first application, at least one slot and the slot parameters of each slot. It can be understood that since the task label is generated based on the first application, when the first task needs to be executed, when the second input of the user is received, the first application can be directly run without the user starting the first application separately, and the slot required for executing the first task and the corresponding slot parameters have been determined. Therefore, when executing the first task, the user does not need to input the slot and the corresponding slot parameters, which further improves the execution efficiency of the first task.

[0083] In the process of generating at least one task tag, the embodiment of the present application determines at least one item of prediction information and the first application required to perform the first task based on the task request information entered by the user in the dialogue input box, further determines at least one slot required to perform the first task and the slot parameters of each slot based on the prediction information and the first application, and finally generates at least one task tag based on the first application, at least one slot, and the slot parameters of each slot. In the process of determining the prediction information, the first application, the slot, and the slot parameters corresponding to the slot, the computing power of the intent prediction model is utilized to not only ensure the accuracy of data determination, but also effectively improve the efficiency of the data determination process.

[0084] In some embodiments of the present application, the intention prediction model generates at least one item of prediction information for the first task based on the task request information, including:

[0085] The intent prediction model extracts task element keywords from task request information;

[0086] The intention prediction model generates at least one item of prediction information for the first task based on historical reference information that matches the task element keywords.

[0087] In some embodiments of the present application, when determining at least one item of prediction information based on a task request in a dialog input box, the intention prediction model may first extract task element keywords from the task request information.

[0088] For example, when a user interacts with an artificial intelligence (AI) assistant, he or she enters the task request "Help me book a ticket" in the dialogue input box of the artificial intelligence (AI) assistant. At this time, the intention prediction model can extract the task element keywords, namely "book" and "ticket", from the information "Help me book a ticket". From the task element keywords, it can be judged that the first task the user needs to perform is the ticket booking task.

[0089] Furthermore, the intention prediction module can generate at least one item of prediction information of the first task based on the historical reference information matched by the task element keywords, thereby predicting the missing task elements of the first task that the user needs to perform.

[0090] For example, after extracting the task element keywords "book" and "ticket" from the information "help me book a ticket", the intention prediction model can preliminarily determine that the first task the user needs to perform is the ticket booking task. At this time, the intention prediction module can match all historical reference information related to ticket booking, and then predict the missing task elements of the ticket booking task that the user needs to perform this time based on the historical reference information related to ticket booking, that is, predict the departure place, destination, travel mode and travel time of the user's ticket booking task based on the historical reference information related to ticket booking.

[0091] The intention prediction model of the embodiment of the present application extracts task element keywords from the task request information input by the user, and then obtains historical reference information that matches the task element keywords. Then, based on the historical reference information, at least one item of prediction information for the first task is generated. In the process of generating the prediction information, the historical reference information that matches the task element keywords can be used as reference data for the prediction information, thereby ensuring that the generated prediction information can match the task request information input by the user, that is, improving the accuracy of the prediction information generation.

[0092] In some embodiments of the present application, the historical reference information includes at least one of the following: historical conversation information of the artificial intelligence AI assistant, device location information, historical purchase information, application usage frequency, and device system parameters;

[0093] The intention prediction model generates at least one item of prediction information for the first task based on historical reference information that matches the task element keywords, including:

[0094] Determine the first task element of the first task based on the historical conversation information of the AI assistant;

[0095] Determine the second task element of the first task based on the device positioning information;

[0096] generating at least one item of prediction information based on the first-category task element and the second-category task element;

[0097] Each item of prediction information includes first-category task elements and second-category task elements.

[0098] In some embodiments of the present application, historical reference information that matches the task element keywords may include historical conversation information of the artificial intelligence AI assistant, device location information, historical purchase information, application usage frequency, device system parameters, etc.

[0099] The AI assistant's historical conversation information can include the context of the conversation information display area on the assistant interface. Specifically, the most recent three rounds of conversation can be obtained based on the maximum unit capacity of the intent prediction model. Alternatively, continuous conversations within 15 minutes can be obtained based on time parameters. Alternatively, historical interaction data related to task request information can be obtained through task element keyword matching. When task element keywords are matched, the historical conversation can be expanded to 5 rounds.

[0100] Frequency of app usage can reflect a user's preference for features when using electronic devices, specifically their tendency to frequently call specific apps or features. Specifically, this can include app selection: for example, prioritizing a particular map app for navigation over other similar tools. Cross-platform service preferences: for example, defaulting to a particular airline or hotel brand when booking a ticket.

[0101] Through historical purchase information, we can understand the user's content preferences and content styles when using electronic devices, specifically the user's personalized needs for information types and service options. The process of obtaining personalized needs includes: data screening logic: for example, giving priority to economy class in air ticket inquiries, and preferring early and late departure times. Preference for technology information in news reading, or preference for entertainment content in short video recommendations. System setting preferences, users' customized settings for device system settings and interface logic. System performance settings: for example, power management preferences for smart power saving mode. Privacy and security settings: for example, when location permissions are disabled, manual address entry is used by default. Notification and sound settings: for example, only important services such as payment reminders are allowed to push strong notifications.

[0102] Device system parameters can include the device information and execution status of the current electronic device, specifically including: (1) the device that the current user is interacting with, such as whether he is operating a mobile phone, a computer, or a watch, because different devices support different capabilities, such as "you are unlikely to run an application on a watch." (2) The applications that the current user's device can operate, so that the large model can know which applications should be used to process user intentions. The applications that can be operated include two categories: third-party applications: such as travel service platforms, taxi applications, map applications, payment applications, etc. System built-in applications: such as phone calls, text messages, address books, settings, alarms, calendars, etc. (3) Some environmental information of the device currently operated by the user, such as The foreground page, whether the ringing status, etc., because some special pages require special recognition, for example: if the user says "zoom in" on the map page, it may mean to zoom in on the map; and if the user says "zoom in" on the desktop, it may mean to zoom in on the desktop icons. Specifically, during the interaction with the input box, the system dynamically captures environmental parameters through the Android system application programming interface (such as ActivityManager to obtain the foreground application), splices them with the user input, and sends them to the first network model to form a multi-modal input of "current device environment + user statement". When the user enters a new message, the locally saved message is injected into the first network model as additional knowledge. The first network model refers to this historical information to determine the user's current intention.

[0103] Furthermore, the process of generating at least one item of prediction information by the intention prediction model based on historical reference information of task element keyword matching may specifically include:

[0104] First, the intent prediction model identifies the first-category task elements for the first task based on the AI assistant's historical conversations. These first-category task elements are those that are missing from the first task. For example, in a ticket booking task, the first-category task element can be a destination element. For example, when interacting with the AI assistant, the user enters the task request "Book me a ticket" into the AI assistant's input box. The intent prediction model can extract the task element keywords "book" and "ticket" from the message "Book me a ticket." Based on these task element keywords, it can be determined that the first task the user needs to perform is ticket booking. The intent prediction model then matches the task element keywords "book" and "ticket" with historical conversations related to ticket booking. From this historical conversation, it determines the destination corresponding to the ticket booking task, which is the first-category task element. For example, if the user repeatedly mentions going to location A in their historical conversations and has previously booked tickets for scenic spot B through the AI assistant, the intent prediction model can determine that the first-category task elements include "location A" and "scenic spot B."

[0105] Furthermore, the intention prediction model can determine the second type of task elements of the first task based on historical purchase information. The second type of task elements are another type of task elements that are missing in executing the first task. In the ticket booking task, the second type of task elements can be the ticket type corresponding to the means of transportation required to reach the destination. For example, after the intention prediction model analyzes the historical conversation information and determines the destination corresponding to the ticket booking task, it can obtain historical purchase information related to ticket booking, that is, determine what types of tickets the user has purchased. For example, the user's historical purchase information includes air tickets to location A, train tickets to location A, and tickets to scenic spot B. At this time, the intention prediction model can determine that the second type of task elements include "airplane", "high-speed rail" and "tickets".

[0106] Finally, the intention prediction model can generate at least one piece of prediction information based on the first and second task elements. The generated prediction information includes both the first and second task elements, that is, the task elements missing from the first task. For example, based on the task element keywords "book" and "zhang", the intention prediction model determines that the first task elements include "place A" and "scenic spot B", and the second task elements include "airplane" and "high-speed rail". Therefore, the corresponding prediction information "air ticket to place A", "high-speed rail to place A", and "ticket to scenic spot B" can be generated.

[0107] The intention prediction model of the embodiment of the present application determines multiple types of task elements of the first task based on historical reference information matched by task element keywords, and then determines at least one prediction information based on the multiple types of task elements, thereby completing the task elements required to execute the first task based on historical reference information matched by task element keywords, and ensuring the accuracy of prediction information generation.

[0108] In some embodiments, Figure 5 Flowchart of the task execution method of some embodiments of the present application; Figure 5 As shown in Figure 2, the process of generating at least one task label by the intent prediction model is as follows:

[0109] Step 302: Generate at least one prediction data of the first task according to the task request information input by the user;

[0110] When the user enters task request information in the dialogue input box of the artificial intelligence AI assistant, the artificial intelligence AI assistant can input the task request information into the intention prediction model. The intention prediction model injects relevant historical reference information based on the task request information, and then generates at least one prediction data related to the task request information based on the task request information and the historical reference information.

[0111] Step 304 , analyzing the first application program that the first task needs to rely on based on the at least one prediction data;

[0112] The intention prediction model can also determine the task that the user needs to perform based on at least one prediction data and the corresponding task request information, such as a ticket booking task. For the task that needs to be performed, the intention prediction model can analyze the first application that can perform the task, and then determine the first application among all applications of the electronic device.

[0113] Step 306, determining the slots that must be filled to execute the first task;

[0114] When executing a task through the first application, it is necessary to determine the slots required to run the first application and the slot parameters that need to be filled in each slot to ensure that the first application can run correctly and thus ensure that the task that the user needs to perform can be completed correctly.

[0115] Step 308, determine whether the slot is filled with the necessary slot parameters; if so, execute step 312, if not, execute step 310;

[0116] After determining the slots required for the operation of the first application, it is also possible to determine whether the necessary slot parameters are filled in each slot. If not, the intent prediction model needs to fill the slots based on the generated prediction data. If all necessary slots are filled with slot parameters, the intent prediction model can generate a task label based on the first application, at least one slot and the slot parameters of each slot.

[0117] Step 310, filling slots according to the predicted data;

[0118] It should be noted that in the process of the intent prediction model determining the predicted data, the first application, at least one slot and the slot parameters of each slot, knowledge injection is required, that is, historical reference information that matches the task element keywords in the task request information is input into the intent prediction model, such as the historical conversations, device information and execution status of the artificial intelligence AI assistant, including the devices interacted by the user, the applications that the device can operate, the environmental information of the device, etc., the user's historical behavior preference library, including function usage preferences, content preferences, system setting preferences, etc., to improve the accuracy of determining the predicted data, the first application, at least one slot and the slot parameters of each slot; through knowledge injection, the slot filling strategy of the status feedback of the electronic device can be integrated to establish a user parameter preference library, and the matching degree of task parameter completion is higher, effectively reducing the input required by the user in specific scenarios.

[0119] Step 312: Generate at least one task tag.

[0120] After the intent prediction model generates at least one task label, all task labels can be displayed in the assistant interface for users to select.

[0121] In some embodiments of the present application, the method further comprises:

[0122] receiving a third input to the assistant interface;

[0123] In response to the third input, displaying a model download interface;

[0124] receiving a fourth input to the model download interface;

[0125] In response to the fourth input, an intent prediction model is downloaded.

[0126] In some embodiments of the present application, before obtaining the task request information input by the user, that is, before predicting the user's intention based on the task request information, the artificial intelligence AI assistant can be operated first to download the intention prediction model, and then the user's intention can be predicted through the intention prediction model.

[0127] Specifically, first, the third input of the user to the assistant interface is received. The third input can be the user's input to the setting control for triggering the display of the model download interface. The third input is used to control the electronic device to display the model download interface. The artificial intelligence AI assistant can display the model download interface in response to the third input. Furthermore, the user can make a fourth input to the model download interface, and the fourth input is used to control the electronic device to download the intention prediction model. In response to the fourth input, the artificial intelligence AI assistant can download the intention prediction model. By downloading the intention prediction model locally, a local intention prediction model of device perception, tool status, and environmental context can be constructed based on the multimodal intention reasoning mechanism of the electronic device itself. Compared with the traditional cloud model prediction method, the user's intention prediction and identification are more accurate.

[0128] For example, Figure 6 This is a schematic diagram of an interface of a task execution method according to some embodiments of the present application; Figure 7 This is a schematic diagram of an interface of a task execution method according to some embodiments of the present application; Figure 8 This is a schematic diagram of the interface of the task execution method of some embodiments of the present application; Figure 6 、 Figure 7 and Figure 8As shown, in the assistant interface of the artificial intelligence AI assistant, a function icon 220 can be displayed. The user can make a third input to the function icon 220. Then, the user clicks "Settings", pulls down to the localized large model function area 230, and clicks to turn on the "Intent Prediction" service. The artificial intelligence AI assistant pops up a window to prompt the user to download the intention prediction local large model interface 224. The "Download Now" control 232 is displayed in the download intention prediction local large model interface 224. The user can make a fourth input to the "Download Now" control 232, so that the "Intent Prediction" service is turned on synchronously after completing the download of the intention prediction model. The above-mentioned third input includes but is not limited to: the user's touch input of the function icon in the assistant interface through a touch device such as a finger or a stylus, or a voice command input by the user, or a specific gesture input by the user, or other feasible input. The specific input can be determined according to actual usage needs and is not limited in the embodiments of this application. The above-mentioned fourth input includes but is not limited to: the user's touch input of the "Download Now" control in the local large model interface for download intention prediction through a touch device such as a finger or a stylus, or a voice command input by the user, or a specific gesture input by the user, or other feasible input, which can be determined according to actual use needs and is not limited in the embodiment of this application. The specific gesture in the embodiment of this application can be any one of a single-click gesture, a sliding gesture, a drag gesture, a pressure recognition gesture, a long press gesture, an area change gesture, a double-press gesture, and a double-click gesture; the click input in the embodiment of this application can be a single-click input, a double-click input, or any number of click inputs, etc., and can also be a long press input or a short press input.

[0129] The embodiment of the present application can download the intention prediction model to the electronic device by operating the assistant interface, so that the user's intention can be predicted based on the task request letter input by the user through the intention prediction model, that is, at least one prediction information of the first task, the first application for executing the first task, at least one slot for executing the first task, and the slot parameters of each slot are generated, and then at least one task label is generated. Compared with the traditional cloud model prediction method, a local intention prediction model of device perception, tool status, and environmental context can be constructed based on the multimodal intention reasoning mechanism of the electronic device itself, so that the user's intention prediction and recognition are more accurate.

[0130] In some embodiments of the present application, in response to the first input, before displaying the first conversation message in the conversation information display area of the assistant interface, the method further includes:

[0131] Receiving a first input to a blank key on an input method keyboard on a dialog input box or an assistant interface;

[0132] In response to the first input, a task tag with the highest prediction accuracy among the at least one task tag is determined as the task tag selected by the user.

[0133] In some embodiments of the present application, before executing the first task, it is first necessary to determine the task tag for executing the first task, that is, first determine the task tag selected by the user, and then determine the task content of the first task, that is, all task elements required for the first task.

[0134] Specifically, before displaying the first conversation message in the conversation information display area of the assistant interface, the user first inputs into the conversation input box or the blank key on the input method keyboard on the assistant interface. Then, in response to the first input, the artificial intelligence (AI) assistant may select the task tag with the highest prediction accuracy among the at least one task tag as the task tag selected by the user.

[0135] For example, Figure 9 Schematic diagram of the interface of the task execution method provided in some embodiments of the present application; Figure 9 As shown, before the first conversation message is displayed in the conversation information display area of the assistant interface, that is, when at least one task label is displayed in the assistant interface, the user can make a first input to the conversation input box 202, or make a first input to the blank key 210 on the input method keyboard on the assistant interface. For example, the user can long press the blank space of the conversation input box 202, or click the blank key 210 on the input method keyboard, and the artificial intelligence AI assistant can directly adopt the leftmost task label 204, which is the task label with the highest prediction accuracy. Then, as Figure 4 As shown, the AI assistant can combine the task request information with the tag content of the task tag selected by the user to generate a first conversation message 220, and display the first conversation message 220 in the conversation information display area of the assistant interface. It should be noted that when displaying at least one task tag, the at least one task tag can be arranged according to the prediction accuracy of the task tag, for example, from the left to the right of the assistant interface in descending order of prediction accuracy.

[0136] The embodiment of the present application receives the user's first input on the blank key on the input method keyboard on the dialogue input box or the assistant interface, and determines the task tag with the highest prediction accuracy among at least one task tag as the task tag selected by the user, thereby facilitating the user to select the task tag with the highest prediction accuracy without the user having to read the label content of each task tag, thereby improving the convenience and efficiency of the user's task tag selection process.

[0137] In some embodiments, the AI assistant can use the intent prediction model to determine the accuracy of each task label. The process of the intent prediction model predicting the user's query intent is mainly based on its autoregressive generation mechanism and context understanding ability. The specific process is as follows:

[0138] First, the input processing phase occurs. During this phase, the task request information entered by the user is first converted into a vector representation through a word embedding layer, with positional encoding applied to preserve word order. The input sequence is processed by a multi-layer translation decoder, which dynamically captures the semantic connections between words in the task request information through a multi-head self-attention mechanism, forming a global contextual representation. The intent prediction model then generates an intent probability distribution: The intent prediction model performs output layer mapping. Specifically, the final layer of the intent prediction model maps the high-dimensional vector to a vocabulary space and uses a softmax function to generate a probability distribution for each possible task element. For example, if the user inputs "help me book a ticket," the output is: {'ticket booking': 90%, 'ticket inquiry': 5%, 'complaint': 5%}. In the intent prediction scenario, the output layer is adapted to directly generate predefined intent labels, such as "ticket booking," "shopping consultation," and "technical support." The intent prediction model then performs autoregressive inference. Specifically, based on the currently processed context, the intent prediction model gradually predicts the next logical task element, ultimately generating a complete intent description or classification label. For deterministic intent prediction tasks, greedy search or beam search is typically used to select the top three output sequences with the highest probability. Finally, the intent prediction model implements a task adaptation mechanism. Specifically, the intent prediction model can be fine-tuned and optimized. Specifically, the pre-trained intent prediction model is fine-tuned on a specific intent recognition dataset. The cross-entropy loss function is used to optimize the model's parameters so that its output probability distribution aligns with the true intent label. For untuned intent prediction models, contextual learning can be performed. By providing a small number of labeled examples in the input, such as "User says 'Book me a ticket' → Ticket Booking," the model is guided to generate intent judgments based on pattern matching.

[0139] The embodiment of the present application determines the prediction accuracy of the task label, thereby ensuring that when the user operates the dialogue input box or the blank key on the input method keyboard, the task label with the highest prediction accuracy can be automatically selected, thereby making it convenient for the user to select the task label with the highest prediction accuracy without the user having to read the label content of each task label, thereby further improving the execution efficiency of the first task.

[0140] In some embodiments of the present application, in response to the first input, before displaying the first conversation message in the conversation information display area of the assistant interface, the method further includes:

[0141] receiving a first input to a task tag of the at least one task tag;

[0142] In response to the first input, the task tag selected by the first input is determined as the task tag selected by the user.

[0143] In some embodiments of the present application, before executing the first task, the user can determine the task tag selected by the user, and then determine the task content of the first task, that is, all task elements required for the first task, based on the task tag selected by the user.

[0144] For example, Figure 10 Schematic diagram of the interface of the task execution method provided in some embodiments of the present application; Figure 10 As shown, before the first conversation message is displayed in the conversation information display area of the assistant interface, that is, when at least one task tag is displayed in the assistant interface, multiple task tags can be expanded in the form of different color tags above the conversation input box 202. The user can complete the selection of the task tag by clicking on any task tag 204 or pressing and holding the task tag 204 with one finger and dragging it to the area where the input method keyboard 226 is located. Then, as shown in FIG. Figure 4 As shown, the artificial intelligence AI assistant can splice the task request information with the tag content of the task tag selected by the user to generate a first dialogue message, and display the first dialogue message in the dialogue information display area of the assistant interface.

[0145] The embodiment of the present application receives the user's first input on one of at least one task tag, thereby enabling the user to select the required task tag by himself, that is, the user determines the task content of the first task, and ensures that the execution of the first task can match the user's intention.

[0146] In some embodiments of the present application, after displaying the first conversation message in the conversation information display area of the assistant interface, the following steps are further included:

[0147] Displaying a task trigger window; the task trigger window includes at least one task trigger card for executing the first task; the task trigger card includes task trigger information determined according to the first conversation message, the task trigger information including all task elements required for the first application to execute the first task and the element content of each task element;

[0148] A second input for at least one task trigger card is received; wherein the second input is used to select the task card.

[0149] In response to the second input, performing the first task includes:

[0150] The first application is controlled to execute the first task according to the task trigger information in the task trigger card selected by the second input.

[0151] In some embodiments of the present application, after the first conversation message is displayed in the conversation information display area of the assistant interface, a task trigger window may also be displayed in the assistant interface. It is understood that after the first conversation message is displayed in the conversation information display area of the assistant interface, it indicates that the user has determined all task elements of the first task to be executed. At this time, upon receiving the user's second input, the first task can be executed. By displaying the task trigger window, the user can be prompted to make the second input at this time, thereby realizing the execution of the first task.

[0152] Specifically, the task trigger window includes at least one task trigger card for executing the first task; the task trigger card includes task trigger information determined based on the first dialogue message, and the task trigger information includes all task elements required by the first application to execute the first task and the element content of each task element.

[0153] Furthermore, the second input received when executing the first task may be the second input executed by the user on the task trigger card. That is, when the second input of the user on the task trigger card is received, the first application required to execute the first task can be controlled to run, so as to execute the first task through the first application according to the task trigger information in the task trigger card selected by the second input.

[0154] In the embodiment of the present application, a task trigger window is displayed, and at least one task trigger card is displayed in the task trigger window. Thus, all task elements and the element content of each task element for executing the first task can be displayed through the at least one task trigger card. This allows the user to select a task trigger card based on the specific element content. After the user selects any task trigger card, the first application can be run based on the element content in the task trigger card, thereby completing the first task.

[0155] In some embodiments of the present application, when the first task is a ticket booking task, displaying a task triggering window includes:

[0156] Displaying a ticket booking task trigger window; the ticket booking task trigger window includes at least one ticket booking task trigger card; the task trigger information in each ticket booking task trigger card includes all ticket booking elements corresponding to a ticket booking option; wherein the ticket booking elements include at least one of the following: departure place, destination, departure time, arrival time, and transportation schedule information;

[0157] Receiving a second input to at least one task trigger card includes:

[0158] receiving a second input for at least one ticket booking task trigger card;

[0159] In response to the second input, performing the first task includes:

[0160] The first application is controlled to execute the ticket booking task according to the ticket booking task trigger information in the ticket booking task trigger card selected by the second input.

[0161] In some embodiments of the present application, when the first task is a ticket booking task, all ticket booking elements corresponding to the ticket booking options may be displayed in the task trigger card. Specifically, the ticket booking elements may include the departure point, destination, departure time, arrival time, and transportation schedule information. By displaying the ticket booking elements, the user can view the departure point, destination, departure time, arrival time, and transportation schedule information during the ticket booking task execution. By displaying multiple task trigger cards, the user can select different ticket booking elements, thereby ensuring the accuracy of the first task execution.

[0162] For example, when a user interacts with the AI assistant, he or she enters "book a ticket for me" in the AI assistant's input box 202. At this time, the AI assistant generates three task tags based on the information "book a ticket for me", namely "air ticket to location A", "high-speed rail to location A" or "ticket to scenic spot B". Then, Figure 11 This is a schematic diagram of the interface of the task execution method of some embodiments of the present application; Figure 11 As shown, the user has selected the task label "Flight to Location A." After the first conversation message is displayed in the dialog information display area of the assistant interface, a booking task trigger window 212 may be displayed in the assistant interface. At least one booking task trigger card 214 for executing the booking task is displayed in booking task trigger window 212. Each booking task trigger card 214 displays all the booking elements required for the booking task, such as the airline name, departure point, destination, departure time, arrival time, transportation schedule, and ticket price. Furthermore, if the same booking element includes different elements, these elements may be displayed via different booking task trigger cards 214, allowing the user to filter the different booking elements and select the desired booking task.

[0163] Furthermore, the second input received when executing the ticket booking task may be the second input executed by the user on the ticket booking task trigger card. That is, when the second input of the user on the ticket booking task trigger card is received, the first application required to execute the ticket booking task can be controlled to run, so as to execute the first task through the first application according to the ticket booking task trigger information in the ticket booking task trigger card selected by the second input.

[0164] For example, Figure 11As shown, each booking task trigger card 214 also displays the words "Go to Book Tickets." The user can click the area where the words "Go to Book Tickets" are located. At this time, the first application will run according to the booking task trigger card selected by the user, including the airline name and flight information, departure point, destination, departure time, arrival time, and ticket price, thereby completing the booking task. Specifically, in one task trigger card, the airline name is M Airlines, the departure point is Location C, the destination is Location A, the departure time is 11:20, the arrival time is 14:00, and the ticket price is 1680 yuan. In another task trigger card, the airline name is N Airlines, the departure point is Location C, the destination is Location A, the departure time is 19:50, the arrival time is 22:25, and the ticket price is 1780 yuan.

[0165] In an embodiment of the present application, when the first task is a ticket booking task, by displaying booking elements such as the departure point, destination, departure time, arrival time, and transportation schedule information in the ticket booking task trigger card in the ticket booking task window, the user can easily view the booking elements. Furthermore, by displaying different booking elements on different ticket booking task trigger cards, the user can select a ticket booking task trigger card based on the different booking elements. After the user selects the ticket booking task trigger card, the first application can be controlled to execute the ticket booking task according to the ticket booking task trigger information in the ticket booking task trigger card selected by the user.

[0166] In some embodiments of the present application, after displaying at least one task label on the assistant interface, the following steps are further included:

[0167] If the first input is not received and the second dialog message is displayed in the dialog information display area, at least one task card is displayed on the assistant interface; the message content of the second dialog message is task request information; each task card includes task trigger information determined based on the second dialog message and historical reference information; the task trigger information includes all task elements required for the first application to perform the second task and the element content of each task element; the second dialog message includes task content describing the second task; the element content of each task element is the element content with the highest prediction accuracy;

[0168] receiving a fifth input from the user regarding the at least one task card;

[0169] In response to the fifth input, the first application is controlled to execute the second task according to the task triggering information in the task card selected by the fifth input.

[0170] In some embodiments of the present application, after at least one task tag is displayed in the assistant interface, if the user does not select any task tag, it means that the prediction information generated by the current artificial intelligence AI assistant is not what the user needs, that is, the tasks that can be executed by all task tags are not the first task that the user needs to perform. At this time, the user can choose to send the task request information entered by the user to the information display area of the assistant interface, that is, display the second dialogue message in the information display area, and the second dialogue message is the task request information entered by the user in the dialogue input box.

[0171] Furthermore, the artificial intelligence AI assistant can make further predictions based on the second dialogue message, thereby displaying at least one task card in the assistant interface. At this time, the task card may include task trigger information determined based on the second dialogue message and the historical reference message, wherein the task trigger information includes all task elements required for the first application to perform the second task and the element content of each task element.

[0172] Furthermore, after displaying at least one task card, a fifth input from the user for at least one task card can be received. The fifth input is used to select a task card. In response to the fifth input, the artificial intelligence AI assistant can control the first application to execute the second task according to the task trigger information in the task card selected by the fifth input.

[0173] For example, Figure 12 Schematic diagram of the interface of the task execution method provided in some embodiments of the present application; Figure 12As shown, task card 216 may display predicted information generated based on the second conversation message. For example, if user 234 sends "Book me a ticket," the AI assistant generates a reply message 236, "What do you need to book?" and displays task card 216. The predicted information displayed in task card 216 is "Book me a ticket to location A." Task card 216 also displays four task elements: time, departure location, service, and destination. These four task elements correspond to the element content "March 15," "Location C," "Airline ticket," and "Location A," respectively. Furthermore, in the task card 216, a "Cancel" option 246 and a "Confirm" option 248 may also be displayed. The user can click the "Confirm" option 248, which is to make a fifth input to the task card 216. The fifth input includes but is not limited to: the user's touch input to the task card 216 through a touch device such as a finger or a stylus, or a voice command input by the user, or a specific gesture input by the user, or other feasible inputs. The specific ones can be determined according to actual usage requirements and are not limited in the embodiments of the present application. In response to the fifth input, the artificial intelligence AI assistant can control the first application to execute the second task according to the task trigger information in the task card 216 selected by the fifth input. Accordingly, the first application is the ticket booking program, which completes the ticket booking task through the ticket booking program in accordance with the task elements and element content in the task card 216, that is, completes the second task.

[0174] Specifically, if Figure 11 As shown, during the execution of the second task, a task trigger window 212 may also be displayed, and task trigger window 212 includes at least one task trigger card 214 for executing the second task. Task trigger card 214 includes task trigger information determined based on the second conversation message, and the task trigger information includes all task elements required for the first application to execute the second task and the element content of each task element. Furthermore, upon receiving a fifth input from the user on task trigger card 214, the first application required for executing the second task may be controlled to run, so that the second task is executed by the first application according to the task trigger information in task trigger card 214 selected by the fifth input.

[0175] In the embodiment of the present application, if the user does not select any task tag, the user can directly send the task request information to the information display area of the assistant interface, that is, display the second dialogue message in the information display area. The artificial intelligence AI assistant can further predict the user's intention based on the second dialogue message and display a task card. The task card may include task trigger information determined based on the second dialogue message and historical reference messages. The task trigger information includes all task elements required for the first application to perform the second task and the element content of each task element. In other words, the second task that the user needs to perform is predicted. When the user makes the fifth input to the task card, the first application can be controlled to perform the second task according to the task trigger information in the task card.

[0176] Furthermore, it is understood that if a user enters a task request in the dialog input box of the assistant interface and does not select any task tags but instead sends the task request, a second dialog message will be displayed in the dialog message display area, and the AI assistant will continue to predict the user's intent based on the second dialog message. During this process, the AI assistant's interface displays the content predicted by the intent prediction model in the form of task cards, rather than directly intruding on the user's conversation with prediction prompts as they are being typed, thus reducing the intrusion on the user's attention.

[0177] In some embodiments of the present application, before displaying at least one task card on the assistant interface, the following steps are further included:

[0178] Inputting the second conversation message into the intent prediction model, whereby the intent prediction model determines the user intent type based on the second conversation message;

[0179] The intention prediction model determines user behavior preference information based on historical purchase information;

[0180] The intent prediction model determines all task elements required to perform the second task and the element content of each task element based on the frequency of occurrence of chat entity words in historical chat dialogue information, user intent type and user behavior preference information.

[0181] In some embodiments of the present application, the AI assistant determines all task elements required for the second task and the content of each task element based on the second conversation message. First, the second conversation message must be input into an intent prediction model. The intent prediction model then determines the user's intent type based on the second conversation message. For example, if the user sends the second conversation message "Book me a ticket," the AI assistant will use the intent prediction model to determine the user's basic intent based on the phrase "Book me a ticket." Here, "book a ticket" may refer to booking a ticket, such as a plane ticket, train ticket, or ticket. Furthermore, the intent prediction model reviews the user's previous conversations with the AI assistant to identify any relevant contextual information that can help understand the specific information of the current request. For example, if the user mentioned information about location A in the previous conversation, or previously inquired about location A's weather, airport, or other information, suggesting potential travel plans, the intent prediction model will prioritize "ticket" over other booking scenarios, such as hotels, based on the "book a ticket" request.

[0182] Furthermore, the intent prediction model can also determine user behavioral preferences based on their past purchase history. Specifically, the intent prediction model accesses the user's behavioral preference database, which contains their past booking records and behavioral patterns. If a user frequently books flights to location A and their preferences, such as time and seat, show a certain pattern, the system can leverage this information to predict their needs. For example, long-term behavioral data might include high-frequency behavior: a user consistently books one to two domestic flights per month, 70% of which are to location A. Time patterns: Monday morning flights are often chosen. Preferences: airline A and economy class are preferred. The large model leverages its reasoning capabilities to infer the probability of the destination: location A has a 70% probability, while other cities have a 30% probability, indicating that location A has the highest probability. Therefore, it defaults to "location A" as the destination and recommends preferred flight times and cabin classes.

[0183] Finally, the intent prediction model can determine all the task elements required to perform the second task and the element content of each task element based on the frequency of occurrence of chat entities in the user's historical chat dialogue information with the AI assistant, the user's intention type, and the user's behavioral preference information.

[0184] Before displaying at least one task card, the embodiment of the present application analyzes the second conversation message input by the user through an intent prediction model to determine the user's intent type, user behavior preference information, and the frequency of occurrence of chat entities in the user's historical chat conversation information with the artificial intelligence (AI) assistant. Furthermore, based on the frequency of occurrence of chat entities in the user's historical chat conversation information with the artificial intelligence (AI) assistant, the user's intent type, and user behavior preference information, all task elements required to execute the second task and the element content of each task element are determined. This ensures the accuracy of all task elements and the element content of each task element displayed in the task card, thereby ensuring that the second task can meet the user's true intent.

[0185] In some embodiments of the present application, the task execution method further includes:

[0186] receiving a sixth input regarding task element content of a first task element in a task card;

[0187] In response to a sixth input, displaying at least two candidate feature content options;

[0188] receiving a selection input of one of the at least two candidate element content options;

[0189] In response to the selection input, the task element content of the first task element is updated to the first element content, where the first element content is the element content corresponding to the alternative element content option selected by the selection input.

[0190] In some embodiments of the present application, for the element content of the task element required to perform the second task displayed in the task card, the user can also modify the element content by operating the task card. Specifically, the user can make a sixth input on the task element content of the first task element in a task card, and the sixth input is used to select the element content that needs to be modified. Upon receiving the sixth input, the artificial intelligence AI assistant can display at least two alternative element content options in response to the sixth input.

[0191] Furthermore, the user can select and input one of the at least two alternative element contents. At this time, the artificial intelligence AI assistant can respond to the user's selection input and update the task element content of the first task element to the first element content selected by the user, thereby completing the modification of the element content of the task element.

[0192] For example, Figure 12 As shown, when the second task is a ticket booking task, four task elements 242 may be displayed in the task card 216, namely time, departure place, service and destination. Correspondingly, for the task element of service, the element content displayed is "air ticket". Figure 14 Schematic diagram of the interface of the task execution method provided in some embodiments of the present application; Figure 14 As shown, at this time, if the user needs to modify the element content, the sixth input can be made to the element content "airplane ticket". Specifically, the sixth input includes but is not limited to: the user's touch input to the element content "airplane ticket" through a touch device such as a finger or a stylus, or a voice command input by the user, or a specific gesture input by the user, or other feasible input. The specific input can be determined according to actual use needs and is not limited in the embodiment of the present application. At this time, at least two alternative element content options 228 can be displayed on one side of the element content. The at least two alternative element content options 228 can be displayed in the form of a list, namely "airplane ticket", "train ticket", and "taxi". Further, the user selects an alternative element content from the at least two alternative element content options 228, that is, the user selects an alternative element. Specifically, the selection input includes but is not limited to: the user's touch input to any alternative element through a touch device such as a finger or a stylus, or a voice command input by the user, or a specific gesture input by the user, or other feasible input. The specific input can be determined according to actual use needs and is not limited in the embodiment of the present application. For example, if the user selects "train ticket", the element content corresponding to the task element "service" can be updated from "airplane ticket" to "train ticket", thereby completing the modification of the element content.

[0193] The embodiment of the present application can display at least two alternative element content options by making a sixth input to the task element content of the first task element in the task card, and then select and input one of the at least two alternative element content options to realize the element content modification of the task element in the task card, that is, the user manually modifies the element content of the task element of the second task to ensure that the second task can meet the user's true intention.

[0194] In some embodiments of the present application, the task execution method further includes:

[0195] receiving a seventh input of task element content of a first task element in a task card;

[0196] In response to the seventh input, the microphone is turned on to receive sound;

[0197] Determine the second factor content based on the user voice information collected by the microphone;

[0198] The task element content of the first task element is updated to the second element content.

[0199] In some embodiments of the present application, for the element content of the task element required to perform the second task displayed in the task card, the user can also modify the element content by voice input. Specifically, the user can make a seventh input on the task element content of the first task element in a task card, and the seventh input is used to select the element content that needs to be modified. When the seventh input is received, the artificial intelligence AI assistant can respond to the seventh input and turn on the microphone to receive sound. When the microphone is turned on to receive sound, the user can perform voice input, that is, the user can say the element content that needs to be modified.

[0200] Furthermore, the AI assistant can determine the second element content based on the user's voice information collected by the microphone, that is, the element content that the user needs to modify. After determining the second element content, the five elements of the first task element can be updated to the second element content, completing the element content modification.

[0201] For example, Figure 12 As shown, in the case where the second task is a ticket booking task, the task card 216 may display four task elements 242, namely time, departure place, service and destination. Correspondingly, for the task element of service, the element content displayed is "air ticket". Figure 15 Schematic diagram of the interface of the task execution method provided in some embodiments of the present application; Figure 15 As shown, if the user needs to modify the element content, he can make a seventh input to the element content of "air ticket", and the seventh input includes but is not limited to: the user uses a touch device such as a finger or a stylus to touch the displayed element content as "air ticket", or a voice command input by the user, or a specific gesture input by the user, or other feasible inputs, which can be determined according to actual use needs and are not limited in the embodiments of the present application. At this time, the microphone recording 218 can be displayed. Furthermore, the user speaks the element content that needs to be modified. For example, the user speaks the voice information of "train ticket", and the artificial intelligence AI assistant can update the element content of the task element "service" from "air ticket" to "train ticket", completing the modification of the element content.

[0202] The embodiment of the present application can display the microphone pickup by making the seventh input to the task element content of the first task element in the task card, and then the microphone can collect the user's voice information, that is, the user can directly say the element content that needs to be modified, and the element content of the task element in the task card can be modified, thereby effectively simplifying the element content modification process, facilitating user operation, and ensuring that the second task can meet the user's true intention.

[0203] In some embodiments of the present application, a task execution method is provided. Figure 16 Flowchart of a task execution method provided for some embodiments of the present application; Figure 16 As shown, the task execution method includes:

[0204] Step 402: Display at least one task card on the assistant interface, while the third conversation message is displayed in the conversation information display area of the assistant interface; wherein the assistant interface is an interface of an artificial intelligence (AI) assistant, and each task card includes prediction information generated based on the third conversation message, the prediction information including missing task elements of the third task; each task card also includes task trigger information determined based on the third conversation message and historical reference information; the task trigger information includes all task elements required for the second application to perform the third task and the element content of each task element; the third conversation message includes task content describing the third task; the element content of each task element is the element content with the highest prediction accuracy;

[0205] In some embodiments of the present application, when a user interacts with an artificial intelligence AI assistant, the user can input a third dialogue message in the dialogue input box of the artificial intelligence AI assistant, and then send the third dialogue message to the dialogue information display area of the assistant interface, so that the third dialogue message is displayed in the dialogue information display area. Specifically, the third dialogue message includes the task content describing the third task. The artificial intelligence AI assistant can make a prediction based on the third dialogue message input by the user, thereby displaying at least one task card in the assistant interface. At this time, each task card includes a prediction information generated based on the third dialogue message, and the prediction information includes the missing task elements of the third task; the task card may include task trigger information determined based on the third dialogue message and the historical reference message, wherein the task trigger message includes all task elements required for the second application to perform the third task and the element content of each task element. In addition, the historical reference information matches the task element keywords extracted from the third dialogue message.

[0206] For example, Figure 13 As shown, the task card may display predicted information generated based on the third conversation message. For example, if the user's third conversation message 238 is "Book me a ticket," the AI assistant generates a reply message 240, "What do you need to book?" and displays a task card 216. The predicted information displayed in task card 216 is "Book me a ticket to location A." Task card 216 also displays four task elements 242: time, departure location, service, and destination. These four task elements correspond to the element content "March 15," "Location C," "Airline ticket," and "Location A," respectively.

[0207] Step 404: receiving a first input from a user on at least one task card;

[0208] Step 406 : Execute a third task in response to the first input.

[0209] In some embodiments of the present application, a user may make a first input to a task card, and in response to the first input, a third task may be executed.

[0210] For example, Figure 13 As shown, task card 216 may also display a "Cancel" option 246 and a "Confirm" option 248. The user can click "Confirm" option 248, thereby making a first input to task card 216. In response to the first input, the AI assistant controls the second application to execute the third task according to the task trigger information in task card 216 selected by the first input. Accordingly, the second application is the ticket booking application. By using the ticket booking application, the user completes the ticket booking task according to the task elements and element content in the task card, thereby completing the third task.

[0211] In an embodiment of the present application, when a third conversation message is displayed in the information display area, the AI assistant can further predict the user's intention based on the third conversation message and display a task card. The task card can include task trigger information determined based on the third conversation message and historical reference messages. The task trigger information includes all task elements required for the second application to perform the third task and the element content of each task element. In other words, the third task required by the user is predicted, and when the user makes a first input on the task card, the second application can be controlled to perform the third task according to the task trigger information in the task card.

[0212] In some embodiments of the present application, before displaying at least one task card on the assistant interface, the following steps are further included:

[0213] Inputting the third conversation message into the intention prediction model, whereby the intention prediction model generates at least one item of prediction information for the third task based on the third conversation message;

[0214] The intent prediction model determines, based on the third conversation message and the at least one item of prediction information, a second application for performing the third task;

[0215] The intention prediction model determines, based on the at least one prediction information and the second application, at least one slot required to perform the third task and slot parameters of each slot;

[0216] The intent prediction model generates at least one task card according to the second application, at least one slot, and slot parameters of each slot.

[0217] In some embodiments of the present application, when the user inputs a third conversation message into the conversation information display area of the assistant interface, the artificial intelligence AI assistant can use the intention prediction model to predict at least one item of prediction information of the third task that the user needs to perform, and use the intention prediction model to determine the second application required to perform the third task.

[0218] Specifically, after the user enters the third dialogue message in the dialogue information display area of the assistant interface, the artificial intelligence AI assistant can first input the third dialogue message into the intention prediction model. The intention prediction model can process the third dialogue message through its own neurons based on the third dialogue message, thereby generating at least one prediction information of the third task, that is, predicting the missing task elements of the third task that the user needs to perform.

[0219] For example, the user sends "help me book a ticket" to the information display area of the artificial intelligence AI assistant. At this time, the artificial intelligence AI assistant can input the information "help me book a ticket" into the intention prediction model. Based on this information, the intention prediction model determines that the user may need to book a plane ticket or train ticket to a certain place, or a ticket to a certain scenic spot. Here, a certain place, plane ticket, train ticket, a certain scenic spot, and ticket are the missing task elements of the first task, that is, the prediction information.

[0220] Furthermore, after determining at least one item of prediction information, the intention prediction model may also determine a second application for performing the third task based on the third conversation message and the at least one item of prediction information.

[0221] Taking the ticket booking task as an example, when the intention prediction model predicts that the third task the user needs to perform is the ticket booking task, it can search for programs that can be used for ticket booking from all applications installed in the electronic device, such as map programs, travel programs, etc., that is, determine the second application required to perform the third task.

[0222] Furthermore, after determining the second application required to execute the third task, the intention prediction model can also determine at least one slot required to execute the third task and slot parameters of each slot based on the prediction information and the corresponding second application.

[0223] Taking the ticket booking task as an example, a ticket booking task requires at least the following slots to satisfy all the task elements: departure location, destination, travel method, and travel time. Accordingly, the corresponding relationship between the slots and their corresponding slot parameters is: departure location: location C, destination: location A, travel method: flight, travel time: March 18th.

[0224] Furthermore, after determining the second application, at least one slot, and the slot parameters of each slot, the intention prediction model can generate at least one task card based on the second application, at least one slot, and the slot parameters of each slot. It is understandable that since the task card is generated based on the second application, when the third task needs to be executed, upon receiving the user's first input, the second application can be run directly without the user having to separately start the second application. In addition, the slot required for executing the third task and the corresponding slot parameters have been determined. Therefore, when executing the third task, the user does not need to input the slot and the corresponding slot parameters, which further improves the execution efficiency of the first task.

[0225] In the process of generating at least one task card, an embodiment of the present application determines at least one item of prediction information and the second application required to perform the third task based on a third dialogue message entered by the user in the dialogue input box. Further, based on the prediction information and the second application, at least one slot required to perform the third task and the slot parameters of each slot are determined. Finally, based on the second application, the at least one slot, and the slot parameters of each slot, at least one task card is generated. In the process of determining the prediction information, the second application, the slot, and the slot parameters corresponding to the slot, utilizing the computing power of the intent prediction model can not only ensure the accuracy of data determination, but also effectively improve the efficiency of the data determination process.

[0226] In some embodiments of the present application, the intent prediction model generates at least one item of prediction information for the third task based on the third conversation message, including:

[0227] The intent prediction model extracts task element keywords from the third conversation message;

[0228] The intention prediction model generates at least one item of prediction information of the third task based on historical reference information that matches the task element keywords.

[0229] In some embodiments of the present application, when determining at least one item of prediction information based on the third dialogue message, the intention prediction model may first extract task element keywords from the third dialogue message.

[0230] For example, when the user is interacting with the artificial intelligence AI assistant, he or she enters the third dialogue message "Help me book a ticket" in the dialogue information display area of the artificial intelligence AI assistant. At this time, the intention prediction model can extract task element keywords from the information "Help me book a ticket", namely "book" and "ticket". From the task element keywords, it can be judged that the third task the user needs to perform is the ticket booking task.

[0231] Furthermore, the intention prediction module can generate at least one prediction information of the third task based on the historical reference information matched by the task element keywords, thereby predicting the missing task elements of the third task that the user needs to perform.

[0232] For example, after extracting the task element keywords "book" and "ticket" from the information "help me book a ticket", the intention prediction model can preliminarily determine that the third task the user needs to perform is the ticket booking task. At this time, the intention prediction module can match all historical reference information related to ticket booking, and then predict the missing task elements of the ticket booking task that the user needs to perform this time based on the historical reference information related to ticket booking, that is, predict the departure place, destination, travel mode and travel time of the user's ticket booking task based on the historical reference information related to ticket booking.

[0233] The intention prediction model in this embodiment of the present application extracts task element keywords from the third conversation message input by the user, obtains historical reference information that matches the task element keywords, and then generates at least one item of prediction information for the third task based on the historical reference information. During the process of generating the prediction information, the historical reference information that matches the task element keywords can be used as reference data for the prediction information, thereby ensuring that the generated prediction information matches the third conversation message input by the user, thereby improving the accuracy of the prediction information generation.

[0234] In some embodiments of the present application, the historical reference information includes at least one of the following: historical conversation information of the artificial intelligence AI assistant, device location information, historical purchase information, application usage frequency, and device system parameters;

[0235] The intention prediction model generates at least one item of prediction information for the third task based on historical reference information that matches the task element keywords, including:

[0236] Determine the first-category task elements of the third task based on the historical conversation information of the AI assistant;

[0237] Determine the second task element of the third task based on the device positioning information;

[0238] generating at least one item of prediction information based on the first-category task element and the second-category task element;

[0239] Each item of prediction information includes first-category task elements and second-category task elements.

[0240] In some embodiments of the present application, historical reference information that matches the task element keywords may include historical conversation information of the artificial intelligence AI assistant, device location information, historical purchase information, application usage frequency, device system parameters, etc.

[0241] Furthermore, the process of generating at least one item of prediction information by the intention prediction model based on historical reference information of task element keyword matching may specifically include:

[0242] First, the intent prediction model determines the first category of task elements for the first task based on the AI assistant's historical conversation information. For example, when a user interacts with the AI assistant, they enter the task request "Help me book a ticket" in the AI assistant's conversation input box. At this time, the intent prediction model can extract the task element keywords "book" and "ticket" from the information "Help me book a ticket." From the task element keywords, it can be determined that the first task the user needs to perform is the ticket booking task. At this time, the intent prediction model matches the historical conversation information related to ticket booking based on the task element keywords "book" and "ticket." From the historical conversation information, it determines the destination corresponding to the ticket booking task, which is the first category of task elements. For example, in the historical conversation information, the user has mentioned going to place A many times, and the user has completed the ticket booking task for scenic spot B in the past through the AI assistant. At this time, the intent prediction model can determine that the first category of task elements includes "place A" and "scenic spot B."

[0243] Furthermore, the intent prediction model can determine the second-category task elements of the first task based on historical purchase information. Taking the ticket booking task as an example, the intent prediction model analyzes historical conversation information to determine the destination corresponding to the ticket booking task. It can then obtain historical purchase information related to ticket booking, that is, determine what types of tickets the user has purchased. For example, if the user's historical purchase information includes a plane ticket to location A, a train ticket to location A, and a ticket to scenic spot B, the intent prediction model can determine that the second-category task elements include "plane," "high-speed rail," and "ticket."

[0244] Finally, the intention prediction model can generate at least one piece of prediction information based on the first and second task elements. The generated prediction information includes both the first and second task elements, that is, the task elements missing from the first task. For example, based on the task element keywords "book" and "zhang", the intention prediction model determines that the first task elements include "place A" and "scenic spot B", and the second task elements include "airplane" and "high-speed rail". Therefore, the corresponding prediction information "air ticket to place A", "high-speed rail to place A", and "ticket to scenic spot B" can be generated.

[0245] The intention prediction model of the embodiment of the present application determines multiple types of task elements of the first task by matching historical reference information based on task element keywords, and then determines at least one prediction information based on the multiple types of task elements, thereby completing the task elements required to execute the first task based on the historical reference information matched based on task element keywords, and improving the accuracy of prediction information generation.

[0246] In some embodiments of the present application, step 406 specifically includes:

[0247] In response to the first input, performing a third task includes:

[0248] Displaying a task trigger window; the task trigger window includes at least one task trigger card for executing the third task; the task trigger card includes task trigger information, and the task trigger information includes all task elements required for the second application to execute the third task and the element content of each task element;

[0249] receiving a second input for at least one task trigger card;

[0250] In response to the second input, the second application is controlled to execute the third task according to the task triggering information in the task triggering card selected by the second input.

[0251] In some embodiments of the present application, upon receiving a first user input on a task card, a task trigger window may be displayed on the assistant interface in response to the first input. It is understood that upon receiving the first user input on the task card, it indicates that the user has determined all task elements of the third task to be executed, and the third task can now be executed. By displaying the task trigger window, the user can be prompted to make a second input, thereby enabling execution of the third task.

[0252] Specifically, the task trigger window includes at least one task trigger card for executing the third task; the task trigger card includes task trigger information determined based on the third dialogue message, and the task trigger information includes all task elements required for the second application to execute the third task and the element content of each task element.

[0253] Furthermore, when the user receives a second input on the task trigger card, the second application required to execute the third task can be controlled to run, so that the third task can be executed through the second application according to the task trigger information in the task trigger card selected by the second input.

[0254] In this embodiment of the present application, a task trigger window is displayed, and at least one task trigger card is displayed in the task trigger window. This allows the at least one task trigger card to display all task elements and the element content of each task element for executing the third task. This allows the user to select a task trigger card based on the specific element content. After selecting any task trigger card, the user can run the second application based on the element content in the task trigger card, thereby completing the third task.

[0255] In some embodiments of the present application, when the third task is a ticket booking task, displaying a task trigger window includes:

[0256] Displaying a ticket booking task trigger window; the ticket booking task trigger window includes at least one ticket booking task trigger card; the task trigger information in each ticket booking task trigger card includes all ticket booking elements corresponding to a ticket booking option; wherein the ticket booking elements include at least one of the following: departure place, destination, departure time, arrival time, and transportation schedule information;

[0257] Receiving a second input to at least one task trigger card includes:

[0258] receiving a second input for at least one ticket booking task trigger card;

[0259] In response to the second input, performing a third task includes:

[0260] The second application is controlled to execute the ticket booking task according to the ticket booking task trigger information in the ticket booking task trigger card selected by the second input.

[0261] In some embodiments of the present application, when the third task is a ticket booking task, all booking elements corresponding to the booking options may be displayed in the task trigger card. Specifically, the booking elements may include the departure point, destination, departure time, arrival time, and transportation schedule information. By displaying the booking elements, the user can view the departure point, destination, departure time, arrival time, and transportation schedule information during the ticket booking task execution. Furthermore, by displaying multiple task trigger cards, the user can select different booking elements, thereby ensuring the accuracy of the first task execution.

[0262] For example, Figure 12 As shown in FIG, when the user interacts with the AI assistant, he inputs “help me order a ticket” in the dialogue information display area of the AI assistant. At this time, the AI assistant generates at least one task card 216 based on the information “help me order a ticket”. Figure 11As shown, after the user makes a first input on a task card, a booking task trigger window 212 may be displayed in the assistant interface. At least one booking task trigger card 214 for executing the booking task is displayed in the booking task trigger window 212. The booking task trigger card 214 displays all the booking elements required for the booking task, such as the airline name, departure point, destination, departure time, arrival time, transportation schedule information, and fares. Furthermore, if the same booking element includes different elements, these elements may be displayed via different booking task trigger cards 214, allowing the user to filter the different booking elements and select the desired booking task.

[0263] Furthermore, the second input received when executing the ticket booking task may be the second input executed by the user on the ticket booking task trigger card. That is, when the second input of the user on the ticket booking task trigger card is received, the second application required to execute the ticket booking task can be controlled to run, so as to execute the third task through the second application according to the ticket booking task trigger information in the ticket booking task trigger card selected by the second input.

[0264] For example, Figure 11 As shown, the words "Go to book tickets" are also displayed in each ticket booking task trigger card 214. The user can click on the area where the words "Go to book tickets" are located. At this time, the first application can run the ticket booking elements such as the airline name, departure place, destination, departure time, arrival time, transportation schedule information, ticket price, etc. in the ticket booking task trigger card selected by the user, and then complete the ticket booking task.

[0265] In this embodiment of the present application, when the third task is a ticket booking task, by displaying booking elements such as the departure point, destination, departure time, arrival time, and transportation schedule information in the ticket booking task trigger card in the ticket booking task window, the user can easily view these booking elements. Furthermore, by displaying different booking elements in different ticket booking task trigger cards, the user can select a ticket booking task trigger card based on the different booking elements. After the user selects the ticket booking task trigger card, the first application is controlled to execute the ticket booking task according to the ticket booking task trigger information in the selected ticket booking task trigger card.

[0266] In some embodiments, the task execution method provided in this application can also be applied to medical consultation tasks. Users can interact with artificial intelligence AI assistants. For example, the user enters "headache for three days" in the dialogue input box of the assistant interface. At this time, the artificial intelligence AI assistant obtains the preliminary symptom information entered by the user, and automatically predicts based on the preliminary symptom information, and then completes the accompanying symptom options, further loads the detection suggestions, and pushes them to the user.

[0267] In some embodiments, the task execution method provided in the present application can also be applied to navigation tasks. The user can interact with the artificial intelligence AI assistant and enter the type of destination. For example, the user enters "find a charging station" in the dialogue input box of the assistant interface. The artificial intelligence AI assistant combines the remaining mileage of the vehicle, the user's preference for the type of destination, etc. to predict the specific destination location for the user, and further recommends a navigation program to the user to complete the navigation task.

[0268] In some embodiments of the present application, a task execution device is provided. Figure 17 A structural block diagram of a task execution device provided in some embodiments of the present application; Figure 17 As shown, the task execution device 1700 includes:

[0269] Display module 1702 is configured to display at least one task tag on the assistant interface when the task request information input by the user is displayed in the dialogue input box of the assistant interface; wherein the assistant interface is an interface of the artificial intelligence (AI) assistant, and a task tag is used to indicate a prediction information generated based on the task request information, and the prediction information includes the task elements missing from the first task;

[0270] Upon receiving the first input, in response to the first input, displaying a first dialogue message in the dialogue information display area of the assistant interface; the first dialogue message comprises task request information and tag content of a task tag selected by the user determined based on the first input; the first dialogue message comprises task content describing the first task, and the task content comprises all task elements for executing the first task;

[0271] The execution module 1704 is configured to execute the first task in response to the second input when the second input is received.

[0272] The task execution device 1700 provided in the embodiment of the present application can predict the actual task elements of the first task based on the task request information when the user enters the task request information in the dialogue input box of the assistant interface, that is, it realizes the prediction of the user's true intention. Furthermore, at least one task label is displayed in the assistant interface to provide the user with at least one result that the user may need for the user to select. When the user's first input is received, the first dialogue message can be displayed in the dialogue information display area to realize the display of all the task elements required to execute the first task. Furthermore, when the user's second input is received, the first task that the user needs to execute can be executed. During the execution of the first task, the user only needs to enter the task request message in the dialogue input box, and the artificial intelligence AI assistant can automatically predict all the task elements of the first task that the user needs to execute, without the user having to input more detailed information or multiple rounds of information, thereby effectively improving the execution efficiency of the electronic device for the first task that the user needs to execute.

[0273] In some embodiments of the present application, the task execution device 1700 further includes:

[0274] an input module, configured to input the task request information into the intention prediction model, and the intention prediction model generates at least one item of prediction information of the first task based on the task request information;

[0275] The intention prediction model determines a first application for performing the first task based on the task request information and the at least one prediction information;

[0276] The intention prediction model determines, based on the at least one prediction information and the first application, at least one slot required to perform the first task and slot parameters of each slot;

[0277] The intent prediction model generates at least one task label according to the first application, at least one slot, and slot parameters of each slot.

[0278] In the process of generating at least one task tag, the embodiment of the present application determines at least one item of prediction information and the first application required to perform the first task based on the task request information entered by the user in the dialogue input box, further determines at least one slot required to perform the first task and the slot parameters of each slot based on the prediction information and the first application, and finally generates at least one task tag based on the first application, at least one slot, and the slot parameters of each slot. In the process of determining the prediction information, the first application, the slot, and the slot parameters corresponding to the slot, the computing power of the intent prediction model is utilized to not only ensure the accuracy of data determination, but also effectively improve the efficiency of the data determination process.

[0279] In some embodiments of the present application, the intention prediction model generates at least one item of prediction information for the first task based on the task request information, including: the intention prediction model extracts task element keywords from the task request information; the intention prediction model generates at least one item of prediction information for the first task based on historical reference information that matches the task element keywords.

[0280] The intention prediction model of the embodiment of the present application extracts task element keywords from the task request information input by the user, and then obtains historical reference information that matches the task element keywords. Then, based on the historical reference information, at least one item of prediction information for the first task is generated. In the process of generating the prediction information, the historical reference information that matches the task element keywords can be used as reference data for the prediction information, thereby ensuring that the generated prediction information can match the task request information input by the user, that is, improving the accuracy of the prediction information generation.

[0281] In some embodiments of the present application, historical reference information includes at least one of the following: historical conversation information of the artificial intelligence AI assistant, device location information, historical purchase information, frequency of application usage, and device system parameters; the intention prediction model is specifically used to: determine the first category of task elements of the first task based on the historical conversation information of the artificial intelligence AI assistant; determine the second category of task elements of the first task based on the device location information; generate at least one prediction information based on the first category of task elements and the second category of task elements; wherein each item of prediction information includes the first category of task elements and the second category of task elements.

[0282] The intention prediction model of the embodiment of the present application determines multiple types of task elements of the first task based on historical reference information matched by task element keywords, and then determines at least one prediction information based on the multiple types of task elements, thereby completing the task elements required to execute the first task based on historical reference information matched by task element keywords, and ensuring the accuracy of prediction information generation.

[0283] In some embodiments of the present application, the receiving module is also used to receive a third input to the assistant interface; the display module 1702 is also used to display the model download interface in response to the third input; the receiving module is also used to receive a fourth input to the model download interface; the execution module 1704 is also used to download the intention prediction model in response to the fourth input.

[0284] The embodiment of the present application can download the intention prediction model to the electronic device by operating the assistant interface, so that the user's intention can be predicted based on the task request letter input by the user through the intention prediction model, that is, at least one prediction information of the first task, the first application for executing the first task, at least one slot for executing the first task, and the slot parameters of each slot are generated, and then at least one task label is generated. Compared with the traditional cloud model prediction method, a local intention prediction model of device perception, tool status, and environmental context can be constructed based on the multimodal intention reasoning mechanism of the electronic device itself, so that the user's intention prediction and recognition are more accurate.

[0285] In some embodiments of the present application, the receiving module is specifically used to receive a first input to the blank key on the input method keyboard on the dialog input box or the assistant interface; the execution module 1704 is specifically used to respond to the first input and determine the task tag with the highest prediction accuracy among at least one task tag as the task tag selected by the user.

[0286] The embodiment of the present application receives the user's first input on the blank key on the input method keyboard on the dialogue input box or the assistant interface, and determines the task tag with the highest prediction accuracy among at least one task tag as the task tag selected by the user, thereby facilitating the user to select the task tag with the highest prediction accuracy without the user having to read the label content of each task tag, thereby improving the convenience and efficiency of the user's task tag selection process.

[0287] In some embodiments of the present application, the receiving module is specifically used to receive a first input of a task tag among at least one task tag; the execution module 1704 is specifically used to determine the task tag selected by the first input as the task tag selected by the user in response to the first input.

[0288] The embodiment of the present application receives the user's first input on one of at least one task tag, thereby enabling the user to select the required task tag by himself, that is, the user determines the task content of the first task, and ensures that the execution of the first task can match the user's intention.

[0289] In some embodiments of the present application, the display module 1702 is also used to: display a task trigger window; the task trigger window includes at least one task trigger card for executing the first task; the task trigger card includes task trigger information determined based on the first dialogue message, and the task trigger information is all task elements required for the first application to execute the first task and the element content of each task element; the receiving module is specifically used to receive a second input for at least one task trigger card; the execution module 1704 is specifically used to control the first application to execute the first task according to the task trigger information in the task trigger card selected by the second input.

[0290] In the embodiment of the present application, a task trigger window is displayed, and at least one task trigger card is displayed in the task trigger window. Thus, all task elements and the element content of each task element for executing the first task can be displayed through the at least one task trigger card. This allows the user to select a task trigger card based on the specific element content. After the user selects any task trigger card, the first application can be run based on the element content in the task trigger card, thereby completing the first task.

[0291] In some embodiments of the present application, when the first task is a ticket booking task, the display module 1702 is specifically used to: display a ticket booking task trigger window; the ticket booking task trigger window includes at least one ticket booking task trigger card; the task trigger information in each ticket booking task trigger card includes all ticket booking elements corresponding to a ticket booking option; wherein the ticket booking elements include at least one of the following: departure place, destination, departure time, arrival time, and transportation schedule information; the receiving module of the task execution device 1700 is specifically used to receive a second input for at least one ticket booking task trigger card; the execution module 1704 is specifically used to control the first application to execute the ticket booking task according to the ticket booking task trigger information in the ticket booking task trigger card selected by the second input.

[0292] In an embodiment of the present application, when the first task is a ticket booking task, by displaying booking elements such as the departure point, destination, departure time, arrival time, and transportation schedule information in the ticket booking task trigger card in the ticket booking task window, the user can easily view the booking elements. Furthermore, by displaying different booking elements on different ticket booking task trigger cards, the user can select a ticket booking task trigger card based on the different booking elements. After the user selects the ticket booking task trigger card, the first application can be controlled to execute the ticket booking task according to the ticket booking task trigger information in the ticket booking task trigger card selected by the user.

[0293] In some embodiments of the present application, the display module 1702 is further configured to:

[0294] When the first input is not received and the second dialogue message is displayed in the dialogue information display area, at least one task card is displayed in the assistant interface; the message content of the second dialogue message is task request information; each task card includes task trigger information determined based on the second dialogue message and historical reference information; the task trigger information includes all task elements required for the first application to perform the second task and the element content of each task element; the second dialogue message includes the task content describing the second task; the element content of each task element is the element content with the highest prediction accuracy; the receiving module of the task execution device 1700 is also used to receive the user's fifth input for at least one task card; the execution module 1704 is also used to control the first application to execute the second task according to the task trigger information in the task card selected by the fifth input in response to the fifth input.

[0295] In the embodiment of the present application, if the user does not select any task tag, the user can directly send the task request information to the information display area of the assistant interface, that is, display the second dialogue message in the information display area. The artificial intelligence AI assistant can further predict the user's intention based on the second dialogue message and display a task card. The task card may include task trigger information determined based on the second dialogue message and historical reference messages. The task trigger information includes all task elements required for the first application to perform the second task and the element content of each task element. In other words, the second task that the user needs to perform is predicted. When the user makes the fifth input to the task card, the first application can be controlled to perform the second task according to the task trigger information in the task card.

[0296] In some embodiments of the present application, the input module is also used to input the second conversation message into the intention prediction model, and the intention prediction model determines the user intention type based on the second conversation message; the intention prediction model determines the user behavior preference information based on the historical purchase information; the intention prediction model determines all task elements required to perform the second task and the element content of each task element based on the frequency of occurrence of chat entity words in the historical chat conversation information, the user intention type and the user behavior preference information.

[0297] Before displaying at least one task card, the embodiment of the present application analyzes the second conversation message input by the user through an intent prediction model to determine the user's intent type, user behavior preference information, and the frequency of occurrence of chat entities in the user's historical chat conversation information with the artificial intelligence (AI) assistant. Furthermore, based on the frequency of occurrence of chat entities in the user's historical chat conversation information with the artificial intelligence (AI) assistant, the user's intent type, and user behavior preference information, all task elements required to execute the second task and the element content of each task element are determined. This ensures the accuracy of all task elements and the element content of each task element displayed in the task card, thereby ensuring that the second task can meet the user's true intent.

[0298] In some embodiments of the present application, the receiving module is also used to receive a sixth input on the task element content of the first task element in a task card; the display module 1702 is also used to display at least two alternative element content options in response to the sixth input; the receiving module is also used to receive a selection input on one of the at least two alternative element content options; the execution module 1704 is also used to update the task element content of the first task element to the first element content in response to the selection input, and the first element content is the element content corresponding to the alternative element content option selected by the selection input.

[0299] The embodiment of the present application can display at least two alternative element content options by making a sixth input to the task element content of the first task element in the task card, and then select and input one of the at least two alternative element content options to realize the element content modification of the task element in the task card, that is, the user manually modifies the element content of the task element of the second task to ensure that the second task can meet the user's true intention.

[0300] In some embodiments of the present application, the receiving module is also used to receive a seventh input of the task element content of the first task element in a task card; the execution module 1704 is also used to turn on the microphone to receive sound in response to the seventh input; determine the second element content based on the user voice information collected by the microphone; and update the task element content of the first task element to the second element content.

[0301] The embodiment of the present application can display the microphone pickup by making the seventh input to the task element content of the first task element in the task card, and then the microphone can collect the user's voice information, that is, the user can directly say the element content that needs to be modified, and the element content of the task element in the task card can be modified, thereby effectively simplifying the element content modification process, facilitating user operation, and ensuring that the second task can meet the user's true intention.

[0302] In some embodiments of the present application, a task execution device 1800 is provided. Figure 18 A structural block diagram of a task execution device provided in some embodiments of the present application; Figure 18 As shown, the task execution device 1800 includes:

[0303] Display module 1802 is configured to display at least one task card on the assistant interface when the third dialogue message is displayed in the dialogue information display area of the assistant interface; wherein the assistant interface is an interface of an artificial intelligence (AI) assistant, and each task card includes prediction information generated based on the third dialogue message, the prediction information including missing task elements of the third task; each task card also includes task trigger information determined based on the third dialogue message and historical reference information; the task trigger information includes all task elements required for the second application to perform the third task and the element content of each task element; the third dialogue message includes task content describing the third task; the element content of each task element is the element content with the highest prediction accuracy;

[0304] Receiving module 1804, configured to receive a first input of a user on at least one task card;

[0305] The execution module 1806 is configured to execute a third task in response to the first input.

[0306] In an embodiment of the present application, when a third conversation message is displayed in the information display area, the AI assistant can further predict the user's intention based on the third conversation message and display a task card. The task card can include task trigger information determined based on the third conversation message and historical reference messages. The task trigger information includes all task elements required for the second application to perform the third task and the element content of each task element. In other words, the third task required by the user is predicted, and when the user makes a first input on the task card, the second application can be controlled to perform the third task according to the task trigger information in the task card.

[0307] In some embodiments of the present application, the task execution device 1800 also includes: an input module for inputting the third dialogue message into the intention prediction model, the intention prediction model generating at least one prediction information of the third task based on the third dialogue message; the intention prediction model determines the second application for executing the third task based on the third dialogue message and the at least one prediction information; the intention prediction model determines at least one slot required to execute the third task and the slot parameters of each slot based on the at least one prediction information and the second application; the intention prediction model generates at least one task label based on the second application, at least one slot and the slot parameters of each slot.

[0308] In the process of generating at least one task card, the embodiment of the present application determines at least one item of prediction information and the second application required to perform the third task based on the task request information entered by the user in the dialogue input box, further determines at least one slot required to perform the third task and the slot parameters of each slot based on the prediction information and the second application, and finally generates at least one task card based on the second application, the at least one slot, and the slot parameters of each slot. In the process of determining the prediction information, the second application, the slot, and the slot parameters corresponding to the slot, the computing power of the intent prediction model is utilized to not only ensure the accuracy of data determination, but also effectively improve the efficiency of the data determination process.

[0309] In some embodiments of the present application, the intention prediction model generates at least one item of prediction information for the third task based on the task request information, including: the intention prediction model extracts task element keywords from the task request information; the intention prediction model generates at least one item of prediction information for the third task based on historical reference information that matches the task element keywords.

[0310] The intention prediction model of the present embodiment extracts task element keywords from the third conversation message input by the user, obtains historical reference information that matches the task element keywords, and then generates at least one item of prediction information for the third task based on the historical reference information. In the process of generating the prediction information, the historical reference information that matches the task element keywords can be used as reference data for the prediction information, thereby ensuring that the generated prediction information matches the task request information input by the user, thereby improving the accuracy of the prediction information generation.

[0311] In some embodiments of the present application, historical reference information includes at least one of the following: historical conversation information of the artificial intelligence AI assistant, device location information, historical purchase information, frequency of use of the application, and device system parameters; the intention prediction model is specifically used to: the intention prediction model generates at least one prediction information of the third task based on the historical reference information that matches the task element keywords, including: determining the first category of task elements of the third task based on the historical conversation information of the artificial intelligence AI assistant; determining the second category of task elements of the third task based on the device location information; generating at least one prediction information based on the first category of task elements and the second category of task elements; wherein each prediction information includes the first category of task elements and the second category of task elements.

[0312] The intention prediction model of the embodiment of the present application determines multiple types of task elements of the first task by matching historical reference information based on task element keywords, and then determines at least one prediction information based on the multiple types of task elements, thereby completing the task elements required to execute the first task based on the historical reference information matched based on task element keywords, and improving the accuracy of prediction information generation.

[0313] In some embodiments of the present application, the display module 1802 is also used to: display a task trigger window; the task trigger window includes at least one task trigger card for executing the third task; the task trigger card includes task trigger information, and the task trigger information includes all task elements required for the second application to execute the third task and the element content of each task element; the receiving module 1804 is specifically used to receive a first input to at least one task trigger card; the execution module 1806 is specifically used to control the second application to execute the third task according to the task trigger information in the task trigger card selected by the first input.

[0314] In this embodiment of the present application, a task trigger window is displayed, and at least one task trigger card is displayed in the task trigger window. This allows the at least one task trigger card to display all task elements and the element content of each task element for executing the third task. This allows the user to select a task trigger card based on the specific element content. After selecting any task trigger card, the user can run the second application based on the element content in the task trigger card, thereby completing the third task.

[0315] In some embodiments of the present application, when the third task is a ticket booking task, the display module 1802 is specifically used to: display a ticket booking task trigger window; the ticket booking task trigger window includes at least one ticket booking task trigger card; the task trigger information in each ticket booking task trigger card includes all ticket booking elements corresponding to a ticket booking option; wherein the ticket booking elements include at least one of the following: departure place, destination, departure time, arrival time, and transportation schedule information; the receiving module 1804 is specifically used to receive a first input for at least one ticket booking task trigger card; the execution module 1806 is specifically used to control the first application to execute the ticket booking task according to the ticket booking task trigger information in the ticket booking task trigger card selected by the first input.

[0316] In this embodiment of the present application, when the third task is a ticket booking task, by displaying booking elements such as the departure point, destination, departure time, arrival time, and transportation schedule information in the ticket booking task trigger card in the ticket booking task window, the user can easily view these booking elements. Furthermore, by displaying different booking elements in different ticket booking task trigger cards, the user can select a ticket booking task trigger card based on the different booking elements. After the user selects the ticket booking task trigger card, the first application is controlled to execute the ticket booking task according to the ticket booking task trigger information in the selected ticket booking task trigger card.

[0317] The control device in the embodiment of the present application can be an electronic device or a component in the electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices other than a terminal. For example, the electronic device can be a mobile phone, a tablet computer, a laptop computer, a PDA, an in-vehicle electronic device, a mobile Internet device (MID), an augmented reality (AR) / virtual reality (VR) device, a robot, a wearable device, an ultra-mobile personal computer (UMPC), a netbook or a personal digital assistant (PDA), etc. It can also be a server, a network attached storage (NAS), a personal computer (PC), a television (TV), a teller machine or a self-service machine, etc., and the embodiment of the present application does not specifically limit it.

[0318] The control device in the embodiment of the present application may be a device having an operating system. The operating system may be an Android operating system, an iOS operating system, or other possible operating systems, which are not specifically limited in the embodiment of the present application.

[0319] The task execution device 1800 provided in the embodiment of the present application can implement each process implemented in the above method embodiment. To avoid repetition, it will not be described here.

[0320] Optionally, an embodiment of the present application further provides an electronic device, Figure 19 A structural block diagram of an electronic device provided in some embodiments of the present application; Figure 19As shown, the electronic device 1900 includes a processor 1902, a memory 1904, and a program or instruction stored in the memory 1904 and executable on the processor 1902. When the program or instruction is executed by the processor 1902, the various processes of the above-mentioned method embodiment are implemented and the same technical effect can be achieved. To avoid repetition, they will not be described here.

[0321] It should be noted that the electronic devices in the embodiments of the present application include the above-mentioned mobile electronic devices and non-mobile electronic devices.

[0322] Figure 20 A schematic diagram of the hardware structure of an electronic device for implementing some embodiments of the present application.

[0323] The electronic device 2000 includes but is not limited to: a radio frequency unit 2001, a network module 2002, an audio output unit 2003, an input unit 2004, a sensor 2005, a display unit 2006, a user input unit 2007, an interface unit 2008, a memory 2009 and a processor 2010 and other components.

[0324] Those skilled in the art will understand that the electronic device 2000 may also include a power source (such as a battery) to supply power to each component, and the power source may be logically connected to the processor 2010 through a power management system, thereby implementing functions such as charging, discharging, and power consumption management through the power management system. Figure 20 The electronic device structure shown in the figure does not constitute a limitation on the electronic device. The electronic device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently, which will not be repeated here.

[0325] Among them, the display unit 2006 is used to display at least one task label on the assistant interface when the task request information input by the user is displayed in the dialogue input box of the assistant interface; wherein the assistant interface is the interface of the artificial intelligence AI assistant, and a task label is used to indicate a prediction information generated based on the task request information, and the prediction information includes the task elements missing from the first task; when a first input is received, in response to the first input, a first dialogue message is displayed in the dialogue information display area of the assistant interface; the first dialogue message is composed of the task request information and the label content of the task label selected by the user determined based on the first input; the first dialogue message includes task content describing the first task, and the task content includes all task elements required to execute the first task; the processor 2010 is used to execute the first task in response to the second input when a second input is received.

[0326] The task execution method provided by the embodiment of the present application can predict the actual task elements of the first task based on the task request information when the user enters the task request information in the dialogue input box of the assistant interface, that is, it realizes the prediction of the user's true intention. Furthermore, at least one task label is displayed in the assistant interface to provide the user with at least one result that the user may need for the user to select. When the user's first input is received, the first dialogue message can be displayed in the dialogue information display area to realize the display of all task elements required to execute the first task. Furthermore, when the user's second input is received, the first task that the user needs to execute can be executed. During the execution of the first task, the user only needs to enter the task request message in the dialogue input box, and the artificial intelligence AI assistant can automatically predict all task elements of the first task that the user needs to execute, without the user having to input more detailed information or multiple rounds of information, thereby effectively improving the execution efficiency of the electronic device for the first task that the user needs to execute.

[0327] In some embodiments of the present application, the processor 2010 is also used to input task request information into the intention prediction model, and the intention prediction model generates at least one prediction information of the first task based on the task request information; the intention prediction model determines the first application for executing the first task based on the task request information and the at least one prediction information; the intention prediction model determines at least one slot required to execute the first task and the slot parameters of each slot based on the at least one prediction information and the first application; the intention prediction model generates at least one task tag based on the first application, the at least one slot and the slot parameters of each slot.

[0328] In the process of generating at least one task tag, the embodiment of the present application determines at least one item of prediction information and the first application required to perform the first task based on the task request information entered by the user in the dialogue input box, further determines at least one slot required to perform the first task and the slot parameters of each slot based on the prediction information and the first application, and finally generates at least one task tag based on the first application, at least one slot, and the slot parameters of each slot. In the process of determining the prediction information, the first application, the slot, and the slot parameters corresponding to the slot, the computing power of the intent prediction model is utilized to not only ensure the accuracy of data determination, but also effectively improve the efficiency of the data determination process.

[0329] In some embodiments of the present application, the intention prediction model generates at least one item of prediction information for the first task based on the task request information, including: the intention prediction model extracts task element keywords from the task request information; the intention prediction model generates at least one item of prediction information for the first task based on historical reference information that matches the task element keywords.

[0330] The intention prediction model of the embodiment of the present application extracts task element keywords from the task request information input by the user, and then obtains historical reference information that matches the task element keywords. Then, based on the historical reference information, at least one item of prediction information for the first task is generated. In the process of generating the prediction information, the historical reference information that matches the task element keywords can be used as reference data for the prediction information, thereby ensuring that the generated prediction information can match the task request information input by the user, that is, improving the accuracy of the prediction information generation.

[0331] In some embodiments of the present application, historical reference information includes at least one of the following: historical conversation information of the artificial intelligence AI assistant, device location information, historical purchase information, frequency of application usage, and device system parameters; the intention prediction model generates at least one prediction information of the first task based on the historical reference information that matches the task element keywords, including: determining the first category of task elements of the first task based on the historical conversation information of the artificial intelligence AI assistant; determining the second category of task elements of the first task based on the device location information; generating at least one prediction information based on the first category of task elements and the second category of task elements; wherein each item of prediction information includes the first category of task elements and the second category of task elements.

[0332] The intention prediction model of the embodiment of the present application determines multiple types of task elements of the first task by matching historical reference information based on task element keywords, and then determines at least one prediction information based on the multiple types of task elements, thereby completing the task elements required to execute the first task based on the historical reference information matched based on task element keywords, and improving the accuracy of prediction information generation.

[0333] In some embodiments of the present application, the user input unit 2007 is also used to receive a third input to the assistant interface; the display unit 2006 is also used to display the model download interface in response to the third input; the user input unit 2007 is also used to receive a fourth input to the model download interface; the processor 2010 is also used to download the intention prediction model in response to the fourth input.

[0334] The embodiment of the present application can download the intention prediction model to the electronic device by operating the assistant interface, so that the user's intention can be predicted according to the task request letter input by the user through the intention prediction model, that is, at least one prediction information of the first task, the first application for executing the first task, at least one slot for executing the first task, and the slot parameters of each slot are generated, and then at least one task label is generated.

[0335] In some embodiments of the present application, the user input unit 2007 is also used to receive a first input to the blank key on the input method keyboard on the dialog input box or the assistant interface; the processor 2010 is also used to determine, in response to the first input, the task tag with the highest prediction accuracy among at least one task tag as the task tag selected by the user.

[0336] The embodiment of the present application receives the user's first input on the blank key on the input method keyboard on the dialogue input box or the assistant interface, and determines the task tag with the highest prediction accuracy among at least one task tag as the task tag selected by the user, thereby facilitating the user to select the task tag with the highest prediction accuracy without the user having to read the label content of each task tag, thereby improving the convenience and efficiency of the user's task tag selection process.

[0337] In some embodiments of the present application, in response to a first input, before displaying a first conversation message in the conversation information display area of the assistant interface, the user input unit 2007 is also used to receive a first input for a task tag among at least one task tag; the processor 2010 is also used to determine, in response to the first input, the task tag selected by the first input as the task tag selected by the user.

[0338] The embodiment of the present application receives the user's first input on one of at least one task tag, thereby enabling the user to select the required task tag by himself, that is, the user determines the task content of the first task, and ensures that the execution of the first task can match the user's intention.

[0339] In some embodiments of the present application, the display unit 2006 is also used to display a task trigger window; the task trigger window includes at least one task trigger card for executing the first task; the task trigger card includes task trigger information determined based on the first dialogue message, and the task trigger information includes all task elements required for the first application to execute the first task and the element content of each task element; the user input unit 2007 is also used to receive a second input to at least one task trigger card; the processor 2010 is specifically used to control the first application to execute the first task according to the task trigger information in the task trigger card selected by the second input.

[0340] In the embodiment of the present application, a task trigger window is displayed, and at least one task trigger card is displayed in the task trigger window. Thus, all task elements and the element content of each task element for executing the first task can be displayed through the at least one task trigger card. This allows the user to select a task trigger card based on the specific element content. After the user selects any task trigger card, the first application can be run based on the element content in the task trigger card, thereby completing the first task.

[0341] In some embodiments of the present application, when the first task is a ticket booking task, the display unit 2006 is specifically used to display a ticket booking task trigger window; the ticket booking task trigger window includes at least one ticket booking task trigger card; the task trigger information in each ticket booking task trigger card includes all ticket booking elements corresponding to a ticket booking option; wherein the ticket booking elements include at least one of the following: departure place, destination, departure time, arrival time, and transportation schedule information; the user input unit is specifically used to receive a second input to at least one ticket booking task trigger card; the processor is specifically used to control the first application to execute the ticket booking task according to the ticket booking task trigger information in the ticket booking task trigger card selected by the second input.

[0342] In an embodiment of the present application, when the first task is a ticket booking task, by displaying booking elements such as the departure point, destination, departure time, arrival time, and transportation schedule information in the ticket booking task trigger card in the ticket booking task window, the user can easily view the booking elements. Furthermore, by displaying different booking elements on different ticket booking task trigger cards, the user can select a ticket booking task trigger card based on the different booking elements. After the user selects the ticket booking task trigger card, the first application can be controlled to execute the ticket booking task according to the ticket booking task trigger information in the ticket booking task trigger card selected by the user.

[0343] In some embodiments of the present application, when the first input is not received and the second dialogue message is displayed in the dialogue information display area, the display unit is also used to display at least one task card on the assistant interface; the message content of the second dialogue message is task request information; each task card includes task trigger information determined based on the second dialogue message and historical reference information; the task trigger information includes all task elements required for the first application to perform the second task and the element content of each task element; the second dialogue message includes task content describing the second task; the element content of each task element is the element content with the highest prediction accuracy; the user input unit 2007 is also used to receive the user's fifth input for at least one task card; the processor is also used to control the first application to perform the second task according to the task trigger information in the task card selected by the fifth input in response to the fifth input.

[0344] In the embodiment of the present application, if the user does not select any task tag, the user can directly send the task request information to the information display area of the assistant interface, that is, display the second dialogue message in the information display area. The artificial intelligence AI assistant can further predict the user's intention based on the second dialogue message and display a task card. The task card may include task trigger information determined based on the second dialogue message and historical reference messages. The task trigger information includes all task elements required for the first application to perform the second task and the element content of each task element. In other words, the second task that the user needs to perform is predicted. When the user makes the fifth input to the task card, the first application can be controlled to perform the second task according to the task trigger information in the task card.

[0345] In some embodiments of the present application, before displaying at least one task card on the assistant interface, the processor 2010 is also used to input the second conversation message into the intention prediction model, and the intention prediction model determines the user intention type based on the second conversation message; the intention prediction model determines the user behavior preference information based on the historical purchase information; the intention prediction model determines all task elements required to perform the second task and the element content of each task element based on the frequency of occurrence of chat entity words in the historical chat conversation information, the user intention type and the user behavior preference information.

[0346] Before displaying at least one task card, the embodiment of the present application analyzes the second conversation message input by the user through an intent prediction model to determine the user's intent type, user behavior preference information, and the frequency of occurrence of chat entities in the user's historical chat conversation information with the artificial intelligence (AI) assistant. Furthermore, based on the frequency of occurrence of chat entities in the user's historical chat conversation information with the artificial intelligence (AI) assistant, the user's intent type, and user behavior preference information, all task elements required to execute the second task and the element content of each task element are determined. This ensures the accuracy of all task elements and the element content of each task element displayed in the task card, thereby ensuring that the second task can meet the user's true intent.

[0347] In some embodiments of the present application, the user input unit 2007 is also used to receive a sixth input on the task element content of the first task element in a task card; the display unit 2006 is also used to display at least two alternative element content options in response to the sixth input; the user input unit 2007 is also used to receive a selection input on one of the at least two alternative element content options; the processor 2010 is also used to update the task element content of the first task element to the first element content in response to the selection input, and the first element content is the element content corresponding to the alternative element content option selected by the selection input.

[0348] The embodiment of the present application can display at least two alternative element content options by making a sixth input to the task element content of the first task element in the task card, and then select and input one of the at least two alternative element content options to realize the element content modification of the task element in the task card, that is, the user manually modifies the element content of the task element of the second task to ensure that the second task can meet the user's true intention.

[0349] In some embodiments of the present application, the user input unit 2007 is also used to receive a seventh input of the task element content of the first task element in a task card; the processor 2010 is also used to turn on the microphone to receive sound in response to the seventh input; determine the second element content based on the user voice information collected by the microphone; and update the task element content of the first task element to the second element content.

[0350] The embodiment of the present application can display the microphone pickup by making the seventh input to the task element content of the first task element in the task card, and then the microphone can collect the user's voice information, that is, the user can directly say the element content that needs to be modified, and the element content of the task element in the task card can be modified, thereby effectively simplifying the element content modification process, facilitating user operation, and ensuring that the second task can meet the user's true intention.

[0351] In some embodiments of the present application, the display unit 2006 is also used to display at least one task card on the assistant interface when the third dialogue message is displayed in the dialogue information display area of the assistant interface; wherein the assistant interface is the interface of an artificial intelligence AI assistant, and each task card includes a prediction information generated based on the third dialogue message, and the prediction information includes the missing task elements of the third task; each task card also includes task trigger information determined based on the third dialogue message and historical reference information; the task trigger information includes all task elements required for the second application to perform the third task and the element content of each task element; the third dialogue message includes the task content describing the third task; the element content of each task element is the element content with the highest prediction accuracy; the user input unit 2007 is also used to receive a first input from the user to at least one task card; the processor 2010 is also used to execute the third task in response to the first input.

[0352] In an embodiment of the present application, when a third conversation message is displayed in the information display area, the AI assistant can further predict the user's intention based on the third conversation message and display a task card. The task card can include task trigger information determined based on the third conversation message and historical reference messages. The task trigger information includes all task elements required for the second application to perform the third task and the element content of each task element. In other words, the third task required by the user is predicted, and when the user makes a first input on the task card, the second application can be controlled to perform the third task according to the task trigger information in the task card.

[0353] In some embodiments of the present application, the processor 2010 is also used to input the third dialogue message into the intention prediction model, and the intention prediction model generates at least one prediction information of the third task based on the third dialogue message; the intention prediction model determines the second application for performing the third task based on the third dialogue message and the at least one prediction information; the intention prediction model determines at least one slot required to perform the third task and the slot parameters of each slot based on the at least one prediction information and the second application; the intention prediction model generates at least one task card based on the second application, at least one slot and the slot parameters of each slot.

[0354] In the process of generating at least one task card, the embodiment of the present application determines at least one item of prediction information and the second application required to perform the third task based on the task request information entered by the user in the dialogue input box, further determines at least one slot required to perform the third task and the slot parameters of each slot based on the prediction information and the second application, and finally generates at least one task card based on the second application, the at least one slot, and the slot parameters of each slot. In the process of determining the prediction information, the second application, the slot, and the slot parameters corresponding to the slot, the computing power of the intent prediction model is utilized to not only ensure the accuracy of data determination, but also effectively improve the efficiency of the data determination process.

[0355] In some embodiments of the present application, the intention prediction model generates at least one item of prediction information for the third task based on the third dialogue message, including: the intention prediction model extracts task element keywords from the third dialogue message; the intention prediction model generates at least one item of prediction information for the third task based on historical reference information that matches the task element keywords.

[0356] The intention prediction model of the present embodiment extracts task element keywords from the third conversation message input by the user, obtains historical reference information that matches the task element keywords, and then generates at least one item of prediction information for the third task based on the historical reference information. In the process of generating the prediction information, the historical reference information that matches the task element keywords can be used as reference data for the prediction information, thereby ensuring that the generated prediction information matches the task request information input by the user, thereby improving the accuracy of the prediction information generation.

[0357] In some embodiments of the present application, historical reference information includes at least one of the following: historical conversation information of the artificial intelligence AI assistant, device location information, historical purchase information, frequency of application usage, and device system parameters; the intention prediction model generates at least one prediction information of the third task based on the historical reference information that matches the task element keywords, including: determining the first category of task elements of the third task based on the historical conversation information of the artificial intelligence AI assistant; determining the second category of task elements of the third task based on the device location information; generating at least one prediction information based on the first category of task elements and the second category of task elements; wherein each item of prediction information includes the first category of task elements and the second category of task elements.

[0358] The intention prediction model of the embodiment of the present application determines multiple types of task elements of the first task by matching historical reference information based on task element keywords, and then determines at least one prediction information based on the multiple types of task elements, thereby completing the task elements required to execute the first task based on the historical reference information matched based on task element keywords, and improving the accuracy of prediction information generation.

[0359] In some embodiments of the present application, the display unit 2006 is specifically used to display a task trigger window; the task trigger window includes at least one task trigger card for executing the third task; the task trigger card includes task trigger information, and the task trigger information includes all task elements required for the second application to execute the third task and the element content of each task element; the user input unit 2007 is specifically used to receive a second input to at least one task trigger card; the processor 2010 is specifically used to respond to the second input, control the second application to execute the third task according to the task trigger information in the task trigger card selected by the second input.

[0360] In this embodiment of the present application, a task trigger window is displayed, and at least one task trigger card is displayed in the task trigger window. This allows the at least one task trigger card to display all task elements and the element content of each task element for executing the third task. This allows the user to select a task trigger card based on the specific element content. After selecting any task trigger card, the user can run the second application based on the element content in the task trigger card, thereby completing the third task.

[0361] In some embodiments of the present application, when the third task is a ticket booking task, the display unit 2006 is specifically used to display a ticket booking task trigger window; the ticket booking task trigger window includes at least one ticket booking task trigger card; the task trigger information in each ticket booking task trigger card includes all ticket booking elements corresponding to a ticket booking option; wherein the ticket booking elements include at least one of the following: departure place, destination, departure time, arrival time, and transportation schedule information; the user input unit 2007 is specifically used to receive a second input to at least one ticket booking task trigger card; the processor 2010 is specifically used to control the second application to execute the ticket booking task according to the ticket booking task trigger information in the ticket booking task trigger card selected by the second input.

[0362] In this embodiment of the present application, when the third task is a ticket booking task, by displaying booking elements such as the departure point, destination, departure time, arrival time, and transportation schedule information in the ticket booking task trigger card in the ticket booking task window, the user can easily view these booking elements. Furthermore, by displaying different booking elements in different ticket booking task trigger cards, the user can select a ticket booking task trigger card based on the different booking elements. After the user selects the ticket booking task trigger card, the first application is controlled to execute the ticket booking task according to the ticket booking task trigger information in the selected ticket booking task trigger card.

[0363] It should be understood that in an embodiment of the present application, the input unit 2004 may include a graphics processing unit (GPU) 20041 and a microphone 20042, and the graphics processing unit 20041 processes the image data of a static image or video obtained by an image capture device (such as a camera) in a video capture mode or an image capture mode. The display unit 2006 may include a display panel 20061, and the display panel 20061 may be configured in the form of a liquid crystal display, an organic light emitting diode, etc. The user input unit 2007 includes a touch panel 20071 and at least one of other input devices 20072. The touch panel 20071 is also called a touch screen. The touch panel 20071 may include two parts: a touch detection device and a touch controller. Other input devices 20072 may include, but are not limited to, a physical keyboard, function keys (such as a volume control button, a switch button, etc.), a trackball, a mouse, and a joystick, which will not be repeated here.

[0364] The memory 2009 can be used to store applications and various data. The memory 2009 may mainly include a first storage area for storing programs or instructions and a second storage area for storing data. The first storage area may store an operating system, applications or instructions required for at least one function (such as a sound playback function, an image playback function, etc.). In addition, the memory 2009 may include a volatile memory or a non-volatile memory, or the memory 2009 may include both volatile and non-volatile memory. The non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct RAM bus random access memory (DRRAM). The memory 2009 in the embodiment of the present application includes but is not limited to these and any other suitable types of memory.

[0365] Processor 2010 may include one or more processing units. Optionally, processor 2010 integrates an application processor and a modem processor. The application processor primarily handles operations related to the operating system, user interface, and application programs, while the modem processor primarily processes wireless communication signals, such as a baseband processor. It is understood that the modem processor may not be integrated into processor 2010.

[0366] An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the various processes of the above-mentioned method embodiment are implemented and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0367] The processor is the processor in the electronic device in the above embodiment. The readable storage medium includes a computer-readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0368] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned method embodiment and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0369] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0370] An embodiment of the present application provides a computer program product, which is stored in a readable storage medium. The program product is executed by at least one processor to implement the various processes of the above-mentioned method embodiment and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0371] It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element. In addition, it should be noted that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in the opposite order according to the functions involved. For example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0372] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a computer software product, which is stored in a readable storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of each embodiment of the present application.

[0373] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are within the protection of this application.

Claims

1. A task execution method, characterized in that: include: When task request information input by a user is displayed in a dialog input box of an assistant interface, at least one task label is displayed on the assistant interface; wherein the assistant interface is an interface of an artificial intelligence (AI) assistant, and a task label is used to indicate a piece of prediction information generated based on the task request information, wherein the prediction information includes a task element missing from the first task; Upon receiving the first input, in response to the first input, displaying a first dialogue message in the dialogue information display area of the assistant interface; the first dialogue message comprises the task request information and label content of a task label selected by the user determined based on the first input; the first dialogue message includes task content describing the first task, and the task content includes all task elements required to perform the first task; In the event that a second input is received, the first task is performed in response to the second input.

2. The task execution method according to claim 1, characterized in that: Before displaying at least one task label on the assistant interface, the method further includes: Inputting the task request information into an intention prediction model, wherein the intention prediction model generates at least one item of prediction information of the first task based on the task request information; The intention prediction model determines a first application for performing the first task based on the task request information and the at least one prediction information; The intention prediction model determines, based on the at least one item of prediction information and the first application, at least one slot required to execute the first task and slot parameters of each slot; The intention prediction model generates at least one task label according to the first application, at least one of the slots, and slot parameters of each slot.

3. The task execution method according to claim 2, characterized in that: The intention prediction model generates at least one item of prediction information for the first task based on the task request information, including: The intention prediction model extracts task element keywords from the task request information; The intention prediction model generates at least one item of prediction information for the first task based on historical reference information that matches the task element keyword.

4. The task execution method according to claim 3, characterized in that: The historical reference information includes at least one of the following: historical conversation information of the artificial intelligence AI assistant, device location information, historical purchase information, application usage frequency, and device system parameters; The intention prediction model generates at least one item of prediction information for the first task based on historical reference information that matches the task element keyword, including: Determining a first category of task elements for the first task based on historical conversation information of the artificial intelligence (AI) assistant; determining a second type of task element of the first task based on the device positioning information; generating at least one item of prediction information based on the first-category task element and the second-category task element; Each item of prediction information includes the first-category task elements and the second-category task elements.

5. The task execution method according to claim 2, characterized in that: The method further comprises: receiving a third input to the assistant interface; In response to the third input, displaying a model download interface; receiving a fourth input to the model download interface; In response to the fourth input, the intent prediction model is downloaded.

6. The task execution method according to claim 3, characterized in that: The method further includes, in response to the first input, before displaying the first conversation message in the conversation information display area of the assistant interface: Receiving a first input to a blank key on the dialog input box or the input method keyboard on the assistant interface; In response to the first input, a task tag with the highest prediction accuracy among the at least one task tag is determined as the task tag selected by the user.

7. The task execution method according to claim 1, characterized in that: The method further includes, in response to the first input, before displaying the first conversation message in the conversation information display area of the assistant interface: receiving a first input to one of the at least one task tag; In response to the first input, the task tag selected by the first input is determined as the task tag selected by the user.

8. The task execution method according to claim 2, characterized in that: After displaying the first conversation message in the conversation information display area of the assistant interface, the method further includes: displaying a task trigger window; the task trigger window including at least one task trigger card for executing the first task; the task trigger card including task trigger information determined based on the first conversation message, the task trigger information including all task elements required for the first application to execute the first task and the element content of each task element; receiving the second input for the at least one task trigger card; In response to the second input, executing the first task includes: The first application is controlled to execute the first task according to the task trigger information in the task trigger card selected by the second input.

9. The task execution method according to claim 8, characterized in that: In the case where the first task is a ticket booking task, displaying the task triggering window includes: Displaying a ticket booking task trigger window; the ticket booking task trigger window includes at least one ticket booking task trigger card; the task trigger information in each ticket booking task trigger card includes all ticket booking elements corresponding to a ticket booking option; wherein the ticket booking elements include at least one of the following: departure place, destination, departure time, arrival time, and transportation schedule information; The receiving the second input of the at least one task trigger card includes: receiving a second input for the at least one ticket booking task trigger card; The performing of the first task in response to the second input includes: Control the first application to execute the ticket booking task according to the ticket booking task trigger information in the ticket booking task trigger card selected by the second input.

10. The task execution method according to claim 1, characterized in that: After displaying at least one task label on the assistant interface, the method further includes: If the first input is not received and the second dialog message is displayed in the dialog information display area, at least one task card is displayed in the assistant interface; the message content of the second dialog message is the task request information; each task card includes task trigger information determined based on the second dialog message and historical reference information; the task trigger information includes all task elements required for the first application to perform the second task and the element content of each task element; the second dialog message includes task content describing the second task; the element content of each task element is the element content with the highest prediction accuracy; receiving a fifth input from the user regarding the at least one task card; In response to the fifth input, the first application is controlled to execute the second task according to the task triggering information in the task card selected by the fifth input.

11. The task execution method according to claim 10, characterized in that: Before displaying at least one task card on the assistant interface, the method further includes: inputting the second conversation message into an intent prediction model, wherein the intent prediction model determines a user intent type based on the second conversation message; The intention prediction model determines user behavior preference information based on historical purchase information; The intention prediction model determines all task elements required to perform the second task and the element content of each task element based on the frequency of occurrence of chat entity words in historical chat dialogue information, the user intention type and the user behavior preference information.

12. The task execution method according to claim 10, characterized in that: The method further comprises: receiving a sixth input regarding task element content of a first task element in a task card; In response to the sixth input, displaying at least two candidate element content options; receiving a selection input of one of the at least two candidate element content options; In response to the selection input, the task element content of the first task element is updated to the first element content, where the first element content is the element content corresponding to the alternative element content option selected by the selection input.

13. The task execution method according to claim 10, characterized in that: The method further comprises: receiving a seventh input of task element content of a first task element in a task card; In response to the seventh input, turning on the microphone to receive sound; Determine the second factor content based on the user voice information collected by the microphone; The task element content of the first task element is updated to the second element content.

14. A task execution method, characterized in that: include: When the third dialogue message is displayed in the dialogue information display area of the assistant interface, at least one task card is displayed on the assistant interface; wherein the assistant interface is an interface of an artificial intelligence (AI) assistant, and each task card includes a prediction information generated based on the third dialogue message, and the prediction information includes the missing task elements of the third task; each task card also includes task trigger information determined based on the third dialogue message and historical reference information; the task trigger information includes all task elements required for the second application to perform the third task and the element content of each task element; the third dialogue message includes the task content describing the third task; and the element content of each task element is the element content with the highest prediction accuracy. receiving a first input from a user regarding the at least one task card; In response to the first input, the third task is performed.

15. A task execution device, characterized in that: include: A display module, configured to display at least one task label on the assistant interface when displaying task request information input by the user in a dialogue input box of the assistant interface; wherein the assistant interface is an interface of an artificial intelligence (AI) assistant, and a task label is used to indicate a piece of prediction information generated based on the task request information, wherein the prediction information includes a task element missing from the first task; Upon receiving the first input, in response to the first input, displaying a first dialogue message in the dialogue information display area of the assistant interface; the first dialogue message comprises the task request information and label content of a task label selected by the user determined based on the first input; the first dialogue message includes task content describing the first task, and the task content includes all task elements required to perform the first task; The execution module is configured to execute the first task in response to the second input when a second input is received.

16. The task execution device according to claim 15, characterized in that: Also includes: an input module, configured to input the task request information into an intention prediction model, wherein the intention prediction model generates at least one item of prediction information of the first task based on the task request information; The intention prediction model determines a first application for performing the first task based on the task request information and the at least one prediction information; The intention prediction model determines, based on the at least one item of prediction information and the first application, at least one slot required to execute the first task and slot parameters of each slot; The intent prediction model generates at least one task label according to the first application, the at least one slot, and slot parameters of each slot.

17. The task execution device according to claim 16, characterized in that: The intention prediction model generates at least one item of prediction information of the first task according to the task request information, including: The intention prediction model extracts task element keywords from the task request information; The intention prediction model generates at least one item of prediction information for the first task based on historical reference information that matches the task element keyword.

18. The task execution device according to claim 17, characterized in that: The historical reference information includes at least one of the following: historical conversation information of the artificial intelligence AI assistant, device location information, historical purchase information, application usage frequency, and device system parameters; The intention prediction model is specifically used for: Determining a first category of task elements for the first task based on historical conversation information of the artificial intelligence (AI) assistant; determining a second type of task element of the first task based on the device positioning information; generating at least one item of prediction information based on the first-category task element and the second-category task element; Each item of prediction information includes the first-category task elements and the second-category task elements.

19. The task execution device according to claim 16, wherein: Also includes: a receiving module, configured to receive a third input to the assistant interface; The display module is further configured to display a model download interface in response to the third input; The receiving module is further configured to receive a fourth input to the model download interface; The execution module is further configured to download the intention prediction model in response to the fourth input.

20. The task execution device according to claim 17, wherein: The receiving module is specifically configured to receive a first input to the dialog input box or the blank key on the input method keyboard on the assistant interface; The execution module is specifically configured to, in response to the first input, determine a task tag with the highest prediction accuracy among the at least one task tag as the task tag selected by the user.

21. The task execution device according to claim 15, characterized in that: The receiving module is specifically configured to receive a first input of a task tag among the at least one task tag; The execution module is specifically configured to, in response to the first input, determine the task tag selected by the first input as the task tag selected by the user.

22. The task execution device according to claim 16, characterized in that: The display module is also used for: displaying a task trigger window; the task trigger window including at least one task trigger card for executing the first task; the task trigger card including task trigger information determined based on the first conversation message, the task trigger information including all task elements required for the first application to execute the first task and the element content of each task element; The receiving module of the task execution device is specifically configured to receive the second input to the at least one task triggering card; The execution module is specifically configured to control the first application to execute the first task according to the task triggering information in the task triggering card selected by the second input.

23. The task execution device according to claim 22, characterized in that: In the case where the first task is a ticket booking task, the display module is specifically configured to: Displaying a ticket booking task trigger window; the ticket booking task trigger window includes at least one ticket booking task trigger card; the task trigger information in each ticket booking task trigger card includes all ticket booking elements corresponding to a ticket booking option; wherein the ticket booking elements include at least one of the following: departure place, destination, departure time, arrival time, and transportation schedule information; The receiving module is specifically configured to receive a second input to the at least one ticket booking task trigger card; The execution module is specifically used to control the first application to execute the ticket booking task according to the ticket booking task trigger information in the ticket booking task trigger card selected by the second input.

24. The task execution device according to claim 15, characterized in that: The display module is also used for: If the first input is not received and the second dialog message is displayed in the dialog information display area, at least one task card is displayed in the assistant interface; the message content of the second dialog message is the task request information; each task card includes task trigger information determined based on the second dialog message and historical reference information; the task trigger information includes all task elements required for the first application to perform the second task and the element content of each task element; the second dialog message includes task content describing the second task; the element content of each task element is the element content with the highest prediction accuracy; The receiving module of the task execution device is further configured to receive a fifth input from the user on the at least one task card; The execution module is further configured to, in response to the fifth input, control the first application to execute the second task according to the task triggering information in the task card selected by the fifth input.

25. The task execution device according to claim 24, characterized in that: The input module is further configured to input the second conversation message into an intention prediction model, wherein the intention prediction model determines a user intention type based on the second conversation message; The intention prediction model determines user behavior preference information based on historical purchase information; The intention prediction model determines all task elements required to perform the second task and the element content of each task element based on the frequency of occurrence of chat entity words in historical chat dialogue information, the user intention type and the user behavior preference information.

26. The task execution device according to claim 24, characterized in that: The receiving module is further configured to receive a sixth input of the task element content of the first task element in a task card; The display module is further configured to display at least two candidate element content options in response to the sixth input; The receiving module is further configured to receive a selection input of one of the at least two candidate element content options; The execution module is further configured to update the task element content of the first task element to a first element content in response to the selection input, where the first element content is the element content corresponding to the alternative element content option selected by the selection input.

27. The task execution device according to claim 24, characterized in that: The receiving module is further configured to receive a seventh input of the task element content of the first task element in a task card; The execution module is further configured to, in response to the seventh input, enable the microphone to receive sound; Determine the second factor content based on the user voice information collected by the microphone; The task element content of the first task element is updated to the second element content.

28. A task execution device, characterized in that: include: A display module, configured to display at least one task card on the assistant interface when the third dialogue message is displayed in the dialogue information display area of the assistant interface; wherein the assistant interface is an interface of an artificial intelligence (AI) assistant, and each task card includes prediction information generated based on the third dialogue message, the prediction information including missing task elements of the third task; each task card also includes task trigger information determined based on the third dialogue message and historical reference information; the task trigger information includes all task elements required for the second application to perform the third task and the element content of each task element; the third dialogue message includes task content describing the third task; the element content of each task element is the element content with the highest prediction accuracy; A receiving module, configured to receive a first input from a user on the at least one task card; An execution module is configured to execute the third task in response to the first input.

29. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the method according to any one of claims 1 to 14 are implemented.

Images